## Problem
The existing behavior isn't exactly incorrect, but is operationally
risky: if the control plane compute hook breaks, then all the control
plane operations trying to call /location_config will end up retrying
forever, which could put more load on the system.
## Summary of changes
- Treat 404s as fatal errors to do fewer retries: a 404 either indicates
we have the wrong URL, or some control plane bug is failing to recognize
our tenant ID as existing.
- Do not return an error on reconcilation errors in a non-creating
/location_config response: this allows the control plane to finish its
Operation (and we will eventually retry the compute notification later)
This PR adds an API to live-reconfigure the VirtualFile io engine.
It also adds a flag to `pagebench get-page-latest-lsn`, which is where I
found this functionality to be useful: it helps compare the io engines
in a benchmark without re-compiling a release build, which took ~50s on
the i3en.3xlarge where I was doing the benchmark.
Switching the IO engine is completely safe at runtime.
## Problem
We have finite amount of runners and intermediate results are often
wanted before a PR is ready for merging. Currently all PRs get e2e tests
run and this creates a lot of throwaway e2e results which may or may not
get to start or complete before a new push.
## Summary of changes
1. Skip e2e test when PR is in draft mode
2. Run e2e when PR status changes from draft to ready for review (change
this to having its trigger in below PR and update results of build and
test)
3. Abstract e2e test in a Separate workflow and call it from the main
workflow for the e2e test
5. Add a label, if that label is present run e2e test in draft
(run-e2e-test-in-draft)
6. Auto add a label(approve to ci) so that all the external contributors
PR , e2e run in draft
7. Document the new label changes and the above behaviour
Draft PR : https://github.com/neondatabase/neon/actions/runs/7729128470
Ready To Review :
https://github.com/neondatabase/neon/actions/runs/7733779916
Draft PR with label :
https://github.com/neondatabase/neon/actions/runs/7725691012/job/21062432342
and https://github.com/neondatabase/neon/actions/runs/7733854028
## Checklist before requesting a review
- [x] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
## Problem
This is mainly to limit our concurrency, rather than to speed up
requests (I was doing some sanity checks on performance of the service
with thousands of shards)
## Summary of changes
- Enable the `diesel:r2d2` feature, which provides an async connection
pool
- Acquire a connection before entering spawn_blocking for a database
transaction (recall that diesel's interface is sync)
- Set a connection pool size of 99 to fit within default postgres limit
(100)
- Also set the tokio blocking thread count to accomodate the same number
of blocking tasks (the only thing we use spawn_blocking for is database
calls).
## Problem
See
https://neondb.slack.com/archives/C04DGM6SMTM/p1707149618314539?thread_ts=1707081520.140049&cid=C04DGM6SMTM
## Summary of changes
Perform checkpoint check after processing `ingest_batch_size` (default
100) WAL records.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
## Problem
We use an outdated version of Python (3.9.2)
## Summary of changes
- Update Python to the latest patch version (3.9.18)
- Unify the usage of python caches where possible
It's awkward to point to a file when doing some kinds of ad-hoc
deployment (like right now, when I'm hacking a helm chart having not
quite hooked up secrets properly yet). We take all the rest of the
secrets as CLI args directly, so let's do the same for public key.
create_neon_superuser runs the first queries in the database after cold
start. Traces suggest that those first queries can make up a significant
fraction of the cold start time. Make it more visible by adding an
explict tracing span to it; currently you just have to deduce it by
looking at the time spent in the parent 'apply_config' span subtracted
by all the other child spans.
## Problem
We've got several issues with the current `benchmarks` job setup:
- `benchmark_durations.json` file (that we generate in runtime to
split tests into several jobs[0]) is not consistent between these
jobs (and very not consistent with the file if we rerun the job). I.e.
test selection for each job can be different, which could end up in
missed tests in a test run.
- `scripts/benchmark_durations` doesn't fetch all tests from the
database (it doesn't expect any extra directories inside
`test_runner/performance`)
- For some reason, currently split into 4 groups ends up with the 4th
group has no tests to run, which fails the job[1]
- [0] https://github.com/neondatabase/neon/pull/4683
- [1] https://github.com/neondatabase/neon/issues/6629
## Summary of changes
- Generate `benchmark_durations.json` file once before we start
`benchmarks` jobs (this makes it consistent across the jobs) and pass
the file content through the GitHub Actions input (this makes it
consistent for reruns)
- `scripts/benchmark_durations` fix SQL query for getting all required
tests
- Split benchmarks into 5 jobs instead of 4 jobs.
When we'll later introduce a global pool of pre-spawned walredo
processes (https://github.com/neondatabase/neon/issues/6581), this
refactoring avoids plumbing through the reference to the pool to all the
places where we create a broken tenant.
Builds atop the refactoring in #6583
Fix several test flakes:
- test_sharding_service_smoke had log failures on "Dropped LSN updates"
- test_emergency_mode had log failures on a deletion queue shutdown
check, where the check was incorrect because it was expecting channel
receiver to stay alive after cancellation token was fired.
- test_secondary_mode_eviction had racing heatmap uploads because the
test was using a live migration hook to set up locations, where that
migration was itself uploading heatmaps and generally making the
situation more complex than it needed to be.
These are the failure modes that I saw when spot checking the last few
failures of each test.
This will mostly/completely address #6511, but I'll leave that ticket
open for a couple days and then check if either of the tests named in
that ticket are flaky.
Related #6511
## Problem
the idea is to keep compute up and running if there are any active
logical replication subscriptions.
### Rationale
Rationale:
- The Write-Ahead Logging (WAL) files, which contain the data changes,
will need to be retained on the publisher side until the subscriber is
able to connect again and apply these changes. This could potentially
lead to increased disk usage on the publisher - and we do not want to
disrupt the source - I think it is more pain for our customer to resolve
storage issues on the source than to pay for the compute at the target.
- Upon resuming the compute resources, the subscriber will start
consuming and applying the changes from the retained WAL files. The time
taken to catch up will depend on the volume of changes and the
configured vCPUs.
we can avoid explaining complex situations where we lag behind (in
extreme cases we could lag behind hours, days or even months)
- I think an important use case for logical replication from a source is
a one-time migration or release upgrade. In this case the customer would
not mind if we are not suspended for the duration of the migration.
We need to document this in the release notes and the documentation in
the context of logical replication where Neon is the target (subscriber)
### See internal discussion here
https://neondb.slack.com/archives/C04DGM6SMTM/p1706793400746539?thread_ts=1706792628.701279&cid=C04DGM6SMTM
Fix cloning the serialized heatmap on every attempt by just turning it
into `bytes::Bytes` before clone so it will be a refcounted instead of
refcounting a vec clone later on.
Also fixes one cancellation token cloning I had missed in #6618.
Cc: #6096
## Problem
We don't have a neat way to carry around migration .sql files during
deploy, and in any case would prefer to avoid depending on diesel CLI to
deploy.
## Summary of changes
- Use `diesel_migrations` crate to embed migrations in our binary
- Run migrations on startup
- Drop the diesel dependency in the `neon_local` binary, as the
attachment_service binary just needs the database to exist. Do database
creation with a simple `createdb`.
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
The solution we ended up for `backoff::retry` requires always cloning of
cancellation tokens even though there is just `.await`. Fix that, and
also turn the return type into `Option<Result<T, E>>` avoiding the need
for the `E::cancelled()` fn passed in.
Cc: #6096
## Problem
This change was left out of #6585 accidentally -- just forgot to push
the very last version of my branch.
Now that we can load database url from Secrets Manager, we don't always
need it on the CLI any more. We should let the user omit it instead of
passing `--database-url ""`
## Summary of changes
- Make `--database-url` optional
Cleanups from https://github.com/neondatabase/neon/pull/6394
- There was a rogue `*` breaking the `GET /tenant/:tenant_id`, which
passes through to shard zero
- There was a duplicate migrate endpoint
- There are un-prefixed API endpoints that were only needed for compat
tests and can now be removed.
## Problem
Running some memory profiling with high concurrent request rate shows
seemingly some memory fragmentation.
## Summary of changes
Eventually, we will want to separate global memory (caches) from local
memory (per connection handshake and per passthrough).
Using a string interner for project info cache helps reduce some of the
fragmentation of the global cache by having a single heap dedicated to
project strings, and not scattering them throughout all a requests.
At the same time, the interned key is 4 bytes vs the 24 bytes that
`SmolStr` offers.
Important: we should only store verified strings in the interner because
there's no way to remove them afterwards. Good for caching responses
from console.
Before tenant migration it made sense to leak broken tenants in the
metrics until restart. Nowdays it makes less sense because on
cancellations we set the tenant broken. The set metric still allows
filterable alerting.
Fixes: #6507
## Problem
We were archiving the pref benchmarks to
- neon DB
- git repo `zenith-perf-data`
As the pref batch ran in parallel when the uploading of results to
zenith-perf-data` git repo resulted in merge conflicts.
Which made the run flaky and as a side effect the build started failing
.
The problem is been expressed in
https://github.com/neondatabase/neon/issues/5160
## Summary of changes
As the results were not used from the git repo it was redundant hence in
this PR cleaning up the results uploading of of perf results to git repo
The shell script `generate_and_push_perf_report.sh` was using a py
script
[git-upload](https://github.com/neondatabase/neon/compare/remove-perf-benchmark-git-upload?expand=1#diff-c6d938e7f060e487367d9dc8055245c82b51a73c1f97956111a495a8a86e9a33)
and
[scripts/generate_perf_report_page.py](https://github.com/neondatabase/neon/pull/6590/files#diff-81af2147e72d07e4cf8ee4395632596d805d6168ba75c71cab58db2659956ef8)
which are not used anywhere else in repo hence also cleaning that up
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat the commit message to not include the
above checklist
There is currently no cleanup done after a delta layer creation error,
so delta layers can accumulate. The problem gets worse as the operation
gets retried and delta layers accumulate on the disk. Therefore, delete
them from disk (if something has been written to disk).
I was getting an error:
/home/heikki/git-sandbox/neon//pgxn/neon_walredo/walredoproc.c:161:5: error: conflicting types for ‘close_range’; have ‘int(unsigned int, unsigned int, unsigned int)’
161 | int close_range(unsigned int start_fd, unsigned int count, unsigned int flags) {
| ^~~~~~~~~~~
In file included from /usr/include/x86_64-linux-gnu/bits/sigstksz.h:24,
from /usr/include/signal.h:328,
from /home/heikki/git-sandbox/neon//pgxn/neon_walredo/walredoproc.c:50:
/usr/include/unistd.h:1208:12: note: previous declaration of ‘close_range’ with type ‘int(unsigned int, unsigned int, int)’
1208 | extern int close_range (unsigned int __fd, unsigned int __max_fd,
| ^~~~~~~~~~~
The discrepancy is in the 3rd argument. Apparently in the glibc
wrapper it's signed.
As a quick fix, rename our close_range() function, the one that calls
syscall() directly, to avoid the clash with the glibc wrapper. In the
long term, an autoconf test would be nice, and some equivalent on
macOS, see issue #6580.
I was on-call this week, these would had made me understand more/faster
of the system:
- move stray attaching start logging inside the span it starts, add
generation
- log ancestor timeline_id or bootstrapping in the beginning of timeline
creation
The issue is still unsolved because of shmem size in VMs. Need to figure it out before applying this patch.
For more details:
```
ERROR: could not resize shared memory segment "/PostgreSQL.2892504480" to 16774205952 bytes: No space left on device
```
As an example, the same issue in community pgvector/pgvector#453.
This includes a compatibility patch that is needed because pgvector
now skips WAL-logging during the index build, and WAL-logs the index
only in one go at the end. That's how GIN, GiST and SP-GIST index
builds work in core PostgreSQL too, but we need some Neon-specific
calls to mark the beginning and end of those build phases.
pgvector is the first index AM that does that with parallel workers,
so I had to modify those functions in the Neon extension to be aware
of parallel workers. Only the leader needs to create the underlying
file and perform the WAL-logging. (In principle, the parallel workers
could participate in the WAL-logging too, but pgvector doesn't do
that. This will need some further work if that changes).
The previous attempt at this (#6592) missed that parallel workers
needed those changes, and segfaulted in parallel build that spilled to
disk.
Testing
-------
We don't have a place for regression tests of extensions at the
moment. I tested this manually with the following script:
```
CREATE EXTENSION IF NOT EXISTS vector;
DROP TABLE IF EXISTS tst;
CREATE TABLE tst (i serial, v vector(3));
INSERT INTO tst (v) SELECT ARRAY[random(), random(), random()] FROM generate_series(1, 15000) g;
-- Serial build, in memory
ALTER TABLE tst SET (parallel_workers=0);
SET maintenance_work_mem='50 MB';
CREATE INDEX idx ON tst USING hnsw (v vector_l2_ops);
-- Test that the index works. (The table contents are random, and the
-- search is approximate anyway, so we cannot check the exact values.
-- For now, just eyeball that they look reasonable)
set enable_seqscan=off;
explain SELECT * FROM tst ORDER BY v <-> ARRAY[0, 0, 0]::vector LIMIT 5;
SELECT * FROM tst ORDER BY v <-> ARRAY[0, 0, 0]::vector LIMIT 5;
DROP INDEX idx;
-- Serial build, spills to on disk
ALTER TABLE tst SET (parallel_workers=0);
SET maintenance_work_mem='5 MB';
CREATE INDEX idx ON tst USING hnsw (v vector_l2_ops);
SELECT * FROM tst ORDER BY v <-> ARRAY[0, 0, 0]::vector LIMIT 5;
DROP INDEX idx;
-- Parallel build, in memory
ALTER TABLE tst SET (parallel_workers=4);
SET maintenance_work_mem='50 MB';
CREATE INDEX idx ON tst USING hnsw (v vector_l2_ops);
SELECT * FROM tst ORDER BY v <-> ARRAY[0, 0, 0]::vector LIMIT 5;
DROP INDEX idx;
-- Parallel build, spills to disk
ALTER TABLE tst SET (parallel_workers=4);
SET maintenance_work_mem='5 MB';
CREATE INDEX idx ON tst USING hnsw (v vector_l2_ops);
SELECT * FROM tst ORDER BY v <-> ARRAY[0, 0, 0]::vector LIMIT 5;
DROP INDEX idx;
```
## Problem
I didn't know about `wait_until` and was relying on `sleep` to wait for
stuff. This caused some tests to be flaky.
https://github.com/neondatabase/neon/issues/6561
## Summary of changes
Switch to `wait_until`, this should make it tests less flaky
At the end of ApplyRecord(), we called pfree on the decoded record, if
it was "oversized". However, we had alread linked it to the "decode
queue" list in XLogReaderState. If we later called XLogBeginRead(), it
called ResetDecoder and tried to free the same record again.
The conditions to hit this are:
- a large WAL record (larger than aboue 64 kB I think, per
DEFAULT_DECODE_BUFFER_SIZE), and
- another WAL record processed by the same WAL redo process after the
large one.
I think the reason we haven't seen this earlier is that you don't get
WAL records that large that are sent to the WAL redo process, except
when logical replication is enabled. Logical replication adds data to
the WAL records, making them larger.
To fix, allocate the buffer ourselves, and don't link it to the decode
queue. Alternatively, we could perhaps have just removed the pfree(),
but frankly I'm a bit scared about the whole queue thing.
## Problem
When we change which physical pageservers a tenant is attached to, we
must update the control plane so that it can update computes. This will
be done via an HTTP hook, as described in
https://www.notion.so/neondatabase/Sharding-Service-Control-Plane-interface-6de56dd310a043bfa5c2f5564fa98365#1fe185a35d6d41f0a54279ac1a41bc94
## Summary of changes
- Optional CLI args `--control-plane-jwt-token` and `-compute-hook-url`
are added. If these are set, then we will use this HTTP endpoint,
instead of trying to use neon_local LocalEnv to update compute
configuration.
- Implement an HTTP-driven version of ComputeHook that calls into the
configured URL
- Notify for all tenants on startup, to ensure that we don't miss
notifications if we crash partway through a change, and carry a
`pending_compute_notification` flag at runtime to allow notifications to
fail without risking never sending the update.
- Add a test for all this
One might wonder: why not do a "forever" retry for compute hook
notifications, rather than carrying a flag on the shard to call
reconcile() again later. The reason is that we will later limit
concurreny of reconciles, when dealing with larger numbers of shards,
and if reconcile is stuck waiting for the control plane to accept a
notification request, it could jam up the whole system and prevent us
making other changes. Anyway: from the perspective of the outside world,
we _do_ retry forever, but we don't retry forever within a given
Reconciler lifetime.
The `pending_compute_notification` logic is predicated on later adding a
background task that just calls `Service::reconcile_all` on a schedule
to make sure that anything+everything that can fail a
Reconciler::reconcile call will eventually be retried.
This test became flaky when postgres retry handling was fixed to use
backoff delays -- each iteration in this test's loop was taking much
longer because pgbench doesn't fail until postgres has given up on
retrying to the pageserver.
We are just removing it, because the condition it tests is no longer
risky: we reload all metadata from remote storage on restart, so
crashing directly between making local changes and doing remote uploads
isn't interesting any more.
Closes: https://github.com/neondatabase/neon/issues/2856
Closes: https://github.com/neondatabase/neon/issues/5329
in `test_statvfs_pressure_{usage,min_avail_bytes}` we now race against
initial logical size calculation on-demand downloading the layers. first
wait out the initial logical sizes, then change the final asserts to be
"eventual", which is not great but it is faster than failing and
retrying.
this issue seems to happen only in debug mode tests.
Fixes: #6510
## Problem
Passing secrets in via CLI/environment is awkward when using helm for
deployment, and not ideal for security (secrets may show up in ps,
/proc).
We can bypass these issues by simply connecting directly to the AWS
Secrets Manager service at runtime.
## Summary of changes
- Add dependency on aws-sdk-secretsmanager
- Update other aws dependencies to latest, to match transitive
dependency versions
- Add `Secrets` type in attachment service, using AWS SDK to load if
secrets are not provided on the command line.
## Problem
not really any problem, just some drive-by changes
## Summary of changes
1. move wake compute
2. move json processing
3. move handle_try_wake
4. move test backend to api provider
5. reduce wake-compute concerns
6. remove duplicate wake-compute loop
## Problem
Sharded tenants only maintain accurate relation sizes on shard 0.
Therefore logical size can only be calculated on shard 0. Fortunately it
is also only _needed_ on shard 0, to provide Safekeeper feedback and to
send consumption metrics.
Closes: #6307
## Summary of changes
- Send 0 for logical size to safekeepers on shards !=0
- Skip logical size warmup task on shards !=0
- Skip imitate_layer_accesses on shards !=0
## Problem
The 5 second activation timeout is appropriate for production
environments, where we want to give a prompt response to the cloud
control plane, and if we fail it will retry the call. In tests however,
we don't want every call to e.g. timeline create to have to come with a
retry wrapper.
This issue has always been there, but it is more apparent in sharding
tests that concurrently attach several tenant shards.
Closes: https://github.com/neondatabase/neon/issues/6563
## Summary of changes
When `testing` feature is enabled, make `ACTIVE_TENANT_TIMEOUT` 30
seconds instead of 5 seconds.
Adds an endpoint to the pageserver to S3-recover an entire tenant to a
specific given timestamp.
Required input parameters:
* `travel_to`: the target timestamp to recover the S3 state to
* `done_if_after`: a timestamp that marks the beginning of the recovery
process. retries of the query should keep this value constant. it *must*
be after `travel_to`, and also after any changes we want to revert, and
must represent a point in time before the endpoint is being called, all
of these time points in terms of the time source used by S3. these
criteria need to hold even in the face of clock differences, so I
recommend waiting a specific amount of time, then taking
`done_if_after`, then waiting some amount of time again, and only then
issuing the request.
Also important to note: the timestamps in S3 work at second accuracy, so
one needs to add generous waits before and after for the process to work
smoothly (at least 2-3 seconds).
We ignore the added test for the mocked S3 for now due to a limitation
in moto: https://github.com/getmoto/moto/issues/7300 .
Part of https://github.com/neondatabase/cloud/issues/8233
## Problem
Follow up to #5461
In my memory usage/fragmentation measurements, these metrics came up as
a large source of small allocations. The replacement metric has been in
use for a long time now so I think it's good to finally remove this.
Per-endpoint data is still tracked elsewhere
## Summary of changes
remove the per-client bytes metrics
A description was written as a follow-on to a section line, rather than
in the proper `description:` part. This caused swagger parsers to
rightly reject it.
This was very useful in debugging the bugs fixed in #6410 and #6502.
There's a lot more we could do. This only adds the printing to delta
layers, not image layers, for example, and it might be useful to print
details of more record types. But this is a good start.
The rust stdlib uses the efficient `posix_spawn` by default.
However, before this PR, pageserver used `pre_exec()` in our
`close_fds()` ext trait.
This PR moves the work that `close_fds()` did to the walredo C code.
I verified manually using `gdb` that we're now forking out the walredo
process using `posix_spawn`.
refs https://github.com/neondatabase/neon/issues/6565
- log when we start walredo process
- include tenant shard id in walredo argv
- dump some basic walredo state in tenant details api
- more suitable walredo process launch histogram buckets
- avoid duplicate tracing labels in walredo launch spans
## Problem
Currently we have no retry mechanism for fetching basebackup. If there's
an unstable connection, starting compute will just fail.
## Summary of changes
Adds an exponential backoff with 7 retries to get the basebackup.
Don't require AWS access keys (AWS_ACCESS_KEY_ID and
AWS_SECRET_ACCESS_KEY) for S3 usage in the pytests, and also allow
AWS_PROFILE to be passed.
One of the two methods is required however.
This allows local development like:
```
aws sso login --profile dev
export ENABLE_REAL_S3_REMOTE_STORAGE=nonempty REMOTE_STORAGE_S3_REGION=eu-central-1 REMOTE_STORAGE_S3_BUCKET=neon-github-ci-tests AWS_PROFILE=dev
cargo build_testing && RUST_BACKTRACE=1 ./scripts/pytest -k debug-pg16 test_runner/regress/test_tenant_delete.py::test_tenant_delete_smoke
```
related earlier PR for the cargo unit tests of the `remote_storage` crate: #6202
---------
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
## Problem
Initially spotted on macOS. When building `attachment_service`, it might
get linked with system `libpq`:
```
$ otool -L target/debug/attachment_service
target/debug/attachment_service:
/opt/homebrew/opt/libpq/lib/libpq.5.dylib (compatibility version 5.0.0, current version 5.16.0)
/System/Library/Frameworks/Security.framework/Versions/A/Security (compatibility version 1.0.0, current version 61040.61.1)
/System/Library/Frameworks/CoreFoundation.framework/Versions/A/CoreFoundation (compatibility version 150.0.0, current version 2202.0.0)
/usr/lib/libiconv.2.dylib (compatibility version 7.0.0, current version 7.0.0)
/usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 1336.61.1)
```
After this PR:
```
$ otool -L target/debug/attachment_service
target/debug/attachment_service:
/Users/bayandin/work/neon/pg_install/v16/lib/libpq.5.dylib (compatibility version 5.0.0, current version 5.16.0)
/System/Library/Frameworks/Security.framework/Versions/A/Security (compatibility version 1.0.0, current version 61040.61.1)
/System/Library/Frameworks/CoreFoundation.framework/Versions/A/CoreFoundation (compatibility version 150.0.0, current version 2202.0.0)
/usr/lib/libiconv.2.dylib (compatibility version 7.0.0, current version 7.0.0)
/usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 1336.61.1)
```
## Summary of changes
- Set `PQ_LIB_DIR` to bundled Postgres 16 lib dir
## Problem
We have switched to new test results and new coverage results, so no
need to collect these data in old formats.
## Summary of changes
- Remove "Upload coverage report" for old coverage report
- Remove "Store Allure test stat in the DB" for old test results format
This commit adds a function to `KeySpace` which updates a key key space
by removing all overlaps with a second key space. This can involve
splitting or removing of existing ranges.
The implementation is not particularly efficient: O(M * N * log(N))
where N is the number of ranges in the current key space and M is the
number of ranges in the key space we are checking against. In practice,
this shouldn't matter much since, in the short term, the only caller of
this function will be the vectored read path and the number of key
spaces invovled will be small. This follows from the upper bound placed
on the number of keys accepted by the vectored read path.
A couple other small utility functions are added. They'll be used by the
vectored search path as well.
When the read path needs to follow a key into the ancestor timeline, it
needs to wait for said ancestor to become active and aware of it's
branching lsn. The logic is lifted into a separate function with it's
own new error type.
This is done because the vectored read path needs the same logic. It's
also the reason for the newly introduced error type.
When we'll switch the read path to proxy into `get_vectored`, we can
remove the duplicated variants from `PageReconstructError`.
This refactoring makes it easier to experimentally replace
BACKGROUND_RUNTIME with a single-threaded runtime. Found this useful
[during benchmarking](https://github.com/neondatabase/neon/pull/6555).
## Problem
Right now if get_role_secret response wasn't cached (e.g. cache already
reached max size) it will send the second (exactly the same request).
## Summary of changes
Avoid needless request.
changes:
- two messages instead of message every second when gate was closing
- replace the gate name string by using a pointer
- slow GateGuards are likely to log who they were (see example)
example found in regress tests: <https://github.com/neondatabase/neon/pull/6542#issuecomment-1919009256>
## Problem
See https://github.com/neondatabase/cloud/issues/8673
## Summary of changes
Download missed SLRU segments from page server
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: Heikki Linnakangas <heikki@neon.tech>
## Problem
The `--path` argument is only used in testing, for compat tests that use
a JSON snapshot of state rather than the postgres database. In regular
deployments, it should be omitted (currently one has to specify `--path
""`)
## Summary of changes
Make `--path` optional.
Some tests which are unit test alike do not need to run on different pg
versions. Logging test is one of them which I found for unrelated
reasons.
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
Depends on: https://github.com/neondatabase/neon/pull/6468
## Problem
The sharding service will be used as a "virtual pageserver" by the
control plane -- so it needs the set of pageserver APIs that the control
plane uses, and to present them under identical URLs, including prefix
(/v1).
## Summary of changes
- Add missing APIs:
- Tenant deletion
- Timeline deletion
- Node list (used in test now, later in tools)
- `/location_config` API (for migrating tenants into the sharding
service)
- Rework attachment service URLs:
- `/v1` prefix is used for pageserver-compatible APIs
- `/upcall/v1` prefix is used for APIs that are called by the pageserver
(re-attach and validate)
- `/debug/v1` prefix is used for endpoints that are for testing
- `/control/v1` prefix is used for new sharding service APIs that do not
mimic a pageserver API, such as registering and configuring nodes.
- Add test_sharding_service. The sharding service already had some
collateral coverage from its use in general tests, but this is the first
dedicated testing for it.
## Problem
See https://neondb.slack.com/archives/C04DGM6SMTM/p1706531433057289
## Summary of changes
1. Do not decrease reconnect timeout until maximal interval value (1
second) is reached
2. Compute reconnect time after connection attempt is taken to exclude
connect time itself from the interval measurement.
So now backend should not perform more than 4 reconnect attempts per
second.
But please notice that backoff is performed locally in each backend and
so if there are many active backends,
then connection (and so error) rate may be much higher.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
## Problem
We currently can't create subscriptions in PG14 and PG15 because only
superusers can, and PG16 requires adding roles to
pg_create_subscription.
## Summary of changes
I added changes to PG14 and PG15 that allow neon_superuser to bypass the
superuser requirement. For PG16, I didn't do that but added a migration
that adds neon_superuser to pg_create_subscription. Also added a test to
make sure it works.
## Problem
PR #6500 has removed the limiting by number of versions/deletions for
time travel calls. We never get informed about how many versions there
are, and thus the call would just hang without any indication of
progress.
## Summary of changes
We improve the pageserver's behaviour with large prefixes, i.e. those
with many keys, removed or currently still available.
* Add a hard limit of 100k versions/deletions. For the reasoning see
https://github.com/neondatabase/cloud/issues/8233#issuecomment-1915021625
, but TLDR it will roughly support tenants of 2 TiB size, of course
depending on general write activity and duration of the s3 retention
window. The goal is to have a limit at all so that the process doesn't
accumulate increasing numbers of versions until an eventual crash.
* Lower the RAM footprint for the `VerOrDelete` datastructure. This
means we now don't cache a lot of redundant metadata in RAM like the
owner ID. The top level datastructure's footprint goes down from 264
bytes to 80 (but it contains strings that are not counted in there).
Follow-up of #6500, part of https://github.com/neondatabase/cloud/issues/8233
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
Since fdatasync is used for flushing WAL, changing file size is unsafe. Make
segment creation atomic by using tmp file + rename to avoid using partially
initialized segments.
fixes https://github.com/neondatabase/neon/issues/6402
It hanged if file size is less than of a normal segment. Normally that doesn't
happen, but it might in case of crash during segment init. We're going to fix
that half initialized segment by durably renaming it after cooking, so this fix
won't be needed, but better avoid busy loop anyway.
fixes https://github.com/neondatabase/neon/issues/6401
Before this patch, when requesting basebackup for a not-found tenant or
timeline, we'd emit an ERROR-level log entry with a huge stack trace.
See #6366 "Details" section for an example
With this patch, we log at INFO level and only a single line.
Example:
```
2024-01-19T14:16:11.479800Z INFO page_service_conn_main{peer_addr=127.0.0.1:43448}: query handler for 'basebackup d69a536d529a68fcf85bc070030cdf4b 035484e9c28d8d0138a492caadd03ffd 0/2204340 --gzip' entity not found: Tenant d69a536d529a68fcf85bc070030cdf4b not found
2024-01-19T14:19:35.807819Z INFO page_service_conn_main{peer_addr=127.0.0.1:48862}: query handler for 'basebackup d69a536d529a68fcf85bc070030cdf4a 035484e9c28d8d0138a492caadd03ffd 0/2204340 --gzip' entity not found: Timeline d69a536d529a68fcf85bc070030cdf4a/035484e9c28d8d0138a492caadd03ffd was not found
```
fixes https://github.com/neondatabase/neon/issues/6366
Changes
-------
- Change `handle_basebackup_request` to return a `QueryError`
- The new `impl From<WaitLsnError> for QueryError` is needed so the `?`
at `wait_lsn()` call in `handle_basebackup_request` works again. It's
duplicating `impl From<WaitLsnError> for PageStreamError`.
- Remove hard-to-spot conversion of `handle_basebackup_request` return
value to anyhow::Result (the place where I replaced `anyhow::Ok` with
`Result::<(), QueryError>::Ok(())`
- Add forgotten distinguished handling for "Tenant not found" case in
`impl From<GetActiveTenantError> for QueryError`
This was not at all pleasant, and I find it very hard to follow the
various error conversions.
It took me a while to spot the hard-to-spot `anyhow::Ok` thing above.
It would have been caught by the compiler if we weren't auto-converting
`anyhow::Error` into `QueryError::Other`.
We should move away from that, in my opinion, instead forcing each
`.context()` site to become `.context().map_err(QueryError::Other)`.
But that's for a future PR.
When using spawn + wait_with_output instead of
std::process::Command::output or tokio::process::Command::output we must
configure the redirection.
Fixes: #6523 by discarding the stdout completely, we only care about
stderr if any.
## Problem
Taking my ideas from https://github.com/neondatabase/neon/pull/6283 and
doing a bit less radical changes. smaller commits.
Proxy flow was quite deeply nested, which makes adding more interesting
error handling quite tricky.
## Summary of changes
I recommend reviewing commit by commit.
1. move handshake logic into a separate file
2. move passthrough logic into a separate file
3. no longer accept a closure in CancelMap session logic
4. Remove connect_to_db, copy logic into handle_client
5. flatten auth_and_wake_compute in authenticate
6. record info for link auth
## Problem
The tenants we want to recover might have tens of thousands of keys, or
more. At that point, the AWS API returns a paginated response.
## Summary of changes
Support paginated responses for `list_object_versions` requests.
Follow-up of #6155, part of https://github.com/neondatabase/cloud/issues/8233
## Problem
Creating sharded tenants will require an instance of the sharding
service -- the initial goal is to deploy one of these in a staging
region (https://github.com/neondatabase/cloud/issues/9718). It will run
as a kubernetes container, similar to the storage broker, so needs to be
built into the container image.
## Summary of changes
Add `attachment_service` binary to container image
## Problem
There's no efficient way of querying the layer map for a range.
## Summary of changes
Introduce a range query for the layer map (`LayerMap::range_search`).
There's two broad steps to it:
1. Find all coverage changes for layers that intersect the queried range
(see `LayerCoverage::range_overlaps`).
The slightly tricky part is dealing with the start of the range. We can
either be aligned with a layer or not and we need
to treat these cases differently.
2. Iterate over the coverage changes and collect the result. For this we
use a two pointer approach: the trailing pointer tracks the start of the
current range (current location in the key space) and the forward
pointer tracks the next coverage change.
Plugging the range search into the read path is deferred to a future PR.
## Performance
I adapted the layer map benchmarks on a local branch. Range searches are
between 2x and 2.5x slower than point searches. That's in line with what I
expected since we query thelayer map twice.
Since `Timeline::get` will proxy to `Timeline::get_vectored` we can
special case the one element layer map range search
at that point.
This is the "partial revert" of #6384. The summaries turned out to be
expensive due to naive vec usage, but also inconclusive because of the
additional context required. In addition to removing summary traces,
small refactoring is done.
## Problem
Measuring cardinality using logs is expensive and slow.
## Summary of changes
Implement a pre-aggregated HyperLogLog-based cardinality estimate.
HyperLogLog estimates the cardinality of a set by using the probability
that the uniform hash of a value will have a run of n 0s at the end is
`1/2^n`, therefore, having observed a run of `n` 0s suggests we have
measured `2^n` distinct values. By using multiple shards, we can use the
harmonic mean to get a more accurate estimate.
We record this into a Prometheus time-series. HyperLogLog counts can be
merged by taking the `max` of each shard. We can apply a `max_over_time`
in order to find the estimate of cardinality of distinct values over
time
## Problem
See https://neondb.slack.com/archives/C06F5UJH601/p1706373716661439
## Summary of changes
Use None instead of 0 as initial accumulator value for calculating
maximal multixact XID.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: Heikki Linnakangas <heikki@neon.tech>
## Problem
http-over-sql allowes host to be in format api.aws.... however it's not
the case for the websocket flow.
## Summary of changes
Relax endpoint check for the ws serverless connections.
PR #5824 introduced the concept of io engines in pageserver and
implemented `tokio-epoll-uring` in addition to our current method,
`std-fs`.
We used `tokio-epoll-uring` in CI for a day to get more exposure to
the code. Now it's time to switch CI back so that we test with `std-fs`
as well, because that's what we're (still) using in production.
## Problem
Spun off from https://github.com/neondatabase/neon/pull/6394 -- this PR
is just the persistence parts and the changes that enable it to work
nicely
## Summary of changes
- Revert #6444 and #6450
- In neon_local, start a vanilla postgres instance for the attachment
service to use.
- Adopt `diesel` crate for database access in attachment service. This
uses raw SQL migrations as the source of truth for the schema, so it's a
soft dependency: we can switch libraries pretty easily.
- Rewrite persistence.rs to use postgres (via diesel) instead of JSON.
- Preserve JSON read+write at startup and shutdown: this enables using
the JSON format in compatibility tests, so that we don't have to commit
to our DB schema yet.
- In neon_local, run database creation + migrations before starting
attachment service
- Run the initial reconciliation in Service::spawn in the background, so
that the pageserver + attachment service don't get stuck waiting for
each other to start, when restarting both together in a test.
The top level retries weren't enough, probably because we do so many
network requests. Fine grained retries ensure that there is higher
potential for the entire test to succeed.
To demonstrate this, consider the following example: let's assume that
each request has 5% chance of failing and we do 10 requests. Then
chances of success without any retries is 0.95^10 = 0.6. With 3 top
level retries it is 1-0.4^3 = 0.936. With 3 fine grained retries it is
(1-0.05^3)^10 = 0.9988 (roundings implicit). So chances of failure are
6.4% for the top level retry vs 0.12% for the fine grained retry.
Follow-up of #6155
## Problem
- `docker.io/neondatabase/build-tools:pinned` image is frequently
outdated on Docker Hub because there's no automated way to update it.
- `update_build_tools_image.yml` workflow contains legacy roll-back
logic, which is not required anymore because it updates only a single
image.
## Summary of changes
- Make `update_build_tools_image.yml` workflow push images to both ECR
and Docker Hub
- Remove unneeded roll-back logic
## Problem
The support for sharding in the pageserver was written before
https://github.com/neondatabase/neon/pull/6205 landed, so when it landed
we couldn't directly test sharding.
## Summary of changes
- Add `test_sharding_smoke` which tests the basics of creating a
sharding tenant, creating a timeline within it, checking that data
within it is distributed.
- Add modes to pg_regress tests for running with 4 shards as well as
with 1.
This patch introduces a new set of grafana metrics for a histogram:
pageserver_get_vectored_seconds_bucket{task_kind="Compaction|PageRequestHandler"}.
While it has a `task_kind` label, only compaction and SLRU fetches are
tracked. This reduces the increase in cardinality to 24.
The metric should allow us to isolate performance regressions while the
vectorized get is being implemented. Once the implementation is
complete, it'll also allow us to quantify the improvements.
## Problem
Triggered `e2e-tests` job is not cancelled along with other jobs in a PR
if the PR get new commits. We can improve the situation by setting
`concurrency_group` for the remote workflow
(https://github.com/neondatabase/cloud/pull/9622 adds
`concurrency_group` group input to the remote workflow).
Ref https://neondb.slack.com/archives/C059ZC138NR/p1706087124297569
Cloud's part added in https://github.com/neondatabase/cloud/pull/9622
## Summary of changes
- Set `concurrency_group` parameter when triggering `e2e-tests`
- At the beginning of a CI pipeline, trigger Cloud's
`cancel-previous-in-concurrency-group.yml` workflow which cancels
previously triggered e2e-tests
refs https://github.com/neondatabase/neon/issues/6473
Before this PR, if process_started() didn't return Ok(true) until we
ran out of retries, we'd return an error but leave the process running.
Try it by adding a 20s sleep to the pageserver `main()`, e.g., right
before we claim the pidfile.
Without this PR, output looks like so:
```
(.venv) cs@devvm-mbp:[~/src/neon-work-2]: ./target/debug/neon_local start
Starting neon broker at 127.0.0.1:50051.
storage_broker started, pid: 2710939
.
attachment_service started, pid: 2710949
Starting pageserver node 1 at '127.0.0.1:64000' in ".neon/pageserver_1".....
pageserver has not started yet, continuing to wait.....
pageserver 1 start failed: pageserver did not start in 10 seconds
No process is holding the pidfile. The process must have already exited. Leave in place to avoid race conditions: ".neon/pageserver_1/pageserver.pid"
No process is holding the pidfile. The process must have already exited. Leave in place to avoid race conditions: ".neon/safekeepers/sk1/safekeeper.pid"
Stopping storage_broker with pid 2710939 immediately.......
storage_broker has not stopped yet, continuing to wait.....
neon broker stop failed: storage_broker with pid 2710939 did not stop in 10 seconds
Stopping attachment_service with pid 2710949 immediately.......
attachment_service has not stopped yet, continuing to wait.....
attachment service stop failed: attachment_service with pid 2710949 did not stop in 10 seconds
```
and we leak the pageserver process
```
(.venv) cs@devvm-mbp:[~/src/neon-work-2]: ps aux | grep pageserver
cs 2710959 0.0 0.2 2377960 47616 pts/4 Sl 14:36 0:00 /home/cs/src/neon-work-2/target/debug/pageserver -D .neon/pageserver_1 -c id=1 -c pg_distrib_dir='/home/cs/src/neon-work-2/pg_install' -c http_auth_type='Trust' -c pg_auth_type='Trust' -c listen_http_addr='127.0.0.1:9898' -c listen_pg_addr='127.0.0.1:64000' -c broker_endpoint='http://127.0.0.1:50051/' -c control_plane_api='http://127.0.0.1:1234/' -c remote_storage={local_path='../local_fs_remote_storage/pageserver'}
```
After this PR, there is no leaked process.
Adds a new `time_travel_recover` function to the `RemoteStorage` trait
that allows time travel like functionality for S3 buckets, regardless of
their content (it is not even pageserver related). It takes a different
approach from [this
post](https://aws.amazon.com/blogs/storage/point-in-time-restore-for-amazon-s3-buckets/)
that is more complicated.
It takes as input a prefix a target timestamp, and a limit timestamp:
* executes [`ListObjectVersions`](https://docs.aws.amazon.com/AmazonS3/latest/API/API_ListObjectVersions.html)
* obtains the latest version that comes before the target timestamp
* copies that latest version to the same prefix
* if there is versions newer than the limit timestamp, it doesn't do
anything for the file
The limit timestamp is meant to be some timestamp before the start of
the recovery operation and after any changes that one wants to revert.
For example, it might be the time point after a tenant was detached from
all involved pageservers. The limiting mechanism ensures that the
operation is idempotent and can be retried without causing additional
writes/copies.
The approach fulfills all the requirements laid out in 8233, and is a
recoverable operation. Nothing is deleted permanently, only new entries
added to the version log.
I also enable [nextest retries](https://nexte.st/book/retries.html) to
help with some general S3 flakiness (on top of low level retries).
Part of https://github.com/neondatabase/cloud/issues/8233
Filter what we log on compaction task. Per discussion in last triage
call, fixing these by introducing and inspecting the root cause within
anyhow::Error instead of rolling out proper conversions.
Fixes: #6365Fixes: #6367
refer #5508
replaces #5837
## Problem
This PR implements sharding support at compute side. Relations are
splinted in stripes and `get_page` requests are redirected to the
particular shard where stripe is located. All other requests (i.e. get
relation or database size) are always send to shard 0.
## Summary of changes
Support of sharding at compute side include three things:
1. Make it possible to specify and change in runtime connection to more
retain one page server
2. Send `get_page` request to the particular shard (determined by hash
of page key)
3. Support multiple servers in prefetch ring requests
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: John Spray <john@neon.tech>
Co-authored-by: Heikki Linnakangas <heikki@neon.tech>
Refactor out test_disk_usage_eviction tenant creation and add a custom
case with 4 tenants, 3 made with pgbench scale=1 and 1 made with pgbench
scale=4.
Because the tenants are created in order of scales [1, 1, 1, 4] this is
simple enough to demonstrate the problem with using absolute access
times, because on a disk usage based eviction run we will
disproportionally target the *first* scale=1 tenant(s), and the later
larger tenant does not lose anything.
This test is not enough to show the difference between `relative_equal`
and `relative_spare` (the fudge factor); much larger scale will be
needed for "the large tenant", but that will make debug mode tests
slower.
Cc: #5304
## Problem
For #6423, creating a reproducer turned out to be very easy, as an
extension to test_ondemand_activation.
However, before I had diagnosed the issue, I was starting with a more
brute force approach of running creation API calls in the background
while restarting a pageserver, and that shows up a bunch of other
interesting issues.
In this PR:
- Add the reproducer for #6423 by extending `test_ondemand_activation`
(confirmed that this test fails if I revert the fix from
https://github.com/neondatabase/neon/pull/6430)
- In timeline creation, return 503 responses when we get an error and
the tenant's cancellation token is set: this covers the cases where we
get an anyhow::Error from something during timeline creation as a result
of shutdown.
- While waiting for tenants to become active during creation, don't
.map_err() the result to a 500: instead let the `From` impl map the
result to something appropriate (this includes mapping shutdown to 503)
- During tenant creation, we were calling `Tenant::load_local` because
no Preload object is provided. This is usually harmless because the
tenant dir is empty, but if there are some half-created timelines in
there, bad things can happen. Propagate the SpawnMode into
Tenant::attach, so that it can properly skip _any_ attempt to load
timelines if creating.
- When we call upsert_location, there's a SpawnMode that tells us
whether to load from remote storage or not. But if the operation is a
retry and we already have the tenant, it is not correct to skip loading
from remote storage: there might be a timeline there. This isn't
strictly a correctness issue as long as the caller behaves correctly
(does not assume that any timelines are persistent until the creation is
acked), but it's a more defensive position.
- If we shut down while the task in Tenant::attach is running, it can
end up spawning rogue tasks. Fix this by holding a GateGuard through
here, and in upsert_location shutting down a tenant after calling
tenant_spawn if we can't insert it into tenants_map. This fixes the
expected behavior that after shutdown_all_tenants returns, no tenant
tasks are running.
- Add `test_create_churn_during_restart`, which runs tenant & timeline
creations across pageserver restarts.
- Update a couple of tests that covered cancellation, to reflect the
cleaner errors we now return.
Makes the `RemoteStorage` trait not be based on `async_trait` any more.
To avoid recursion in async (not supported by Rust), we made
`GenericRemoteStorage` generic on the "Unreliable" variant. That allows
us to have the unreliable wrapper never contain/call itself.
related earlier work: #6305
## Problem
too many string based IDs. easy to mix up ID types.
## Summary of changes
Add a bunch of `SmolStr` wrappers that provide convenience methods but
are type safe
Fixes: #6459 by formatting full causes of an error to log, while keeping
the top level string for end-user.
Changes user visible error detail from:
```
-DETAIL: page server returned error: Read error: Failed to reconstruct a page image:
+DETAIL: page server returned error: Read error
```
However on pageserver logs:
```
-ERROR page_service_conn_main{...}: error reading relation or page version: Read error: Failed to reconstruct a page image:
+ERROR page_service_conn_main{...}: error reading relation or page version: Read error: reconstruct a page image: launch walredo process: spawn process: Permission denied (os error 13)
```
## Problem
The initdb cancellation added in #5921 is not sufficient to reliably
abort the entire initdb process. Initdb also spawns children. The tests
added by #6310 (#6385) and #6436 now do initdb cancellations on a more
regular basis.
In #6385, I attempted to issue `killpg` (after giving it a new process
group ID) to kill not just the initdb but all its spawned subprocesses,
but this didn't work. Initdb doesn't take *that* long in the end either,
so we just wait until it concludes.
## Summary of changes
* revert initdb cancellation support added in #5921
* still return `Err(Cancelled)` upon cancellation, but this is just to
not have to remove the cancellation infrastructure
* fixes to the `test_tenant_delete_races_timeline_creation` test to make
it reliably pass
Fixes#6385
## Problem
The API for detaching things wasn't implement yet, but one could hit
this case indirectly from tests when using attach-hook, and find tenants
unexpectedly attached again because their policy remained Single.
## Summary of changes
Add PlacementPolicy::Detached, and:
- add the behavior for it in schedule()
- in tenant_migrate, refuse if the policy is detached
- automatically set this policy in attach-hook if the caller has
specified pageserver=null.
The pagebench integration PR (#6214) issues attachment requests in
parallel.
We observed corrupted attachments.json from time to time, especially in
the test cases with high tenant counts.
The atomic overwrite added in #6444 exposed the root cause cleanly:
the `.commit()` calls of two request handlers could interleave or
be reordered.
See also:
https://github.com/neondatabase/neon/pull/6444#issuecomment-1906392259
This PR makes changes to the `persistence` module to fix above race:
- mpsc queue for PendingWrites
- one writer task performs the writes in mpsc queue order
- request handlers that need to do writes do it using the
new `mutating_transaction` function.
`mutating_transaction`, while holding the lock, does the modifications,
serializes the post-modification state, and pushes that as a
`PendingWrite` into the mpsc queue.
It then release the lock and `await`s the completion of the write.
The writer tasks executes the `PendingWrites` in queue order.
Once the write has been executed, it wakes the writing tokio task.
The pagebench integration PR (#6214) is the first to SIGQUIT & then
restart attachment_service.
With many tenants (100), we have found frequent failures on restart in
the CI[^1].
[^1]:
[Allure](https://neon-github-public-dev.s3.amazonaws.com/reports/pr-6214/7615750160/index.html#suites/e26265675583c610f99af77084ae58f1/851ff709578c4452/)
```
2024-01-22T19:07:57.932021Z INFO request{method=POST path=/attach-hook request_id=2697503c-7b3e-4529-b8c1-d12ef912d3eb}: Request handled, status: 200 OK
2024-01-22T19:07:58.898213Z INFO Got SIGQUIT. Terminating
2024-01-22T19:08:02.176588Z INFO version: git-env:d56f31639356ed8e8ce832097f132f27ee19ac8a, launch_timestamp: 2024-01-22 19:08:02.174634554 UTC, build_tag build_tag-env:7615750160, state at /tmp/test_output/test_pageserver_max_throughput_getpage_at_latest_lsn[10-13-30]/repo/attachments.json, listening on 127.0.0.1:15048
thread 'main' panicked at /__w/neon/neon/control_plane/attachment_service/src/persistence.rs:95:17:
Failed to load state from '/tmp/test_output/test_pageserver_max_throughput_getpage_at_latest_lsn[10-13-30]/repo/attachments.json': trailing characters at line 1 column 8957 (maybe your .neon/ dir was written by an older version?)
stack backtrace:
0: rust_begin_unwind
at /rustc/82e1608dfa6e0b5569232559e3d385fea5a93112/library/std/src/panicking.rs:645:5
1: core::panicking::panic_fmt
at /rustc/82e1608dfa6e0b5569232559e3d385fea5a93112/library/core/src/panicking.rs:72:14
2: attachment_service::persistence::PersistentState::load_or_new::{{closure}}
at ./control_plane/attachment_service/src/persistence.rs:95:17
3: attachment_service::persistence::Persistence:🆕:{{closure}}
at ./control_plane/attachment_service/src/persistence.rs:103:56
4: attachment_service::main::{{closure}}
at ./control_plane/attachment_service/src/main.rs:69:61
5: tokio::runtime::park::CachedParkThread::block_on::{{closure}}
at ./.cargo/registry/src/index.crates.io-6f17d22bba15001f/tokio-1.34.0/src/runtime/park.rs:282:63
6: tokio::runtime::coop::with_budget
at ./.cargo/registry/src/index.crates.io-6f17d22bba15001f/tokio-1.34.0/src/runtime/coop.rs:107:5
7: tokio::runtime::coop::budget
at ./.cargo/registry/src/index.crates.io-6f17d22bba15001f/tokio-1.34.0/src/runtime/coop.rs:73:5
8: tokio::runtime::park::CachedParkThread::block_on
at ./.cargo/registry/src/index.crates.io-6f17d22bba15001f/tokio-1.34.0/src/runtime/park.rs:282:31
9: tokio::runtime::context::blocking::BlockingRegionGuard::block_on
at ./.cargo/registry/src/index.crates.io-6f17d22bba15001f/tokio-1.34.0/src/runtime/context/blocking.rs:66:9
10: tokio::runtime::scheduler::multi_thread::MultiThread::block_on::{{closure}}
at ./.cargo/registry/src/index.crates.io-6f17d22bba15001f/tokio-1.34.0/src/runtime/scheduler/multi_thread/mod.rs:87:13
11: tokio::runtime::context::runtime::enter_runtime
at ./.cargo/registry/src/index.crates.io-6f17d22bba15001f/tokio-1.34.0/src/runtime/context/runtime.rs:65:16
12: tokio::runtime::scheduler::multi_thread::MultiThread::block_on
at ./.cargo/registry/src/index.crates.io-6f17d22bba15001f/tokio-1.34.0/src/runtime/scheduler/multi_thread/mod.rs:86:9
13: tokio::runtime::runtime::Runtime::block_on
at ./.cargo/registry/src/index.crates.io-6f17d22bba15001f/tokio-1.34.0/src/runtime/runtime.rs:350:50
14: attachment_service::main
at ./control_plane/attachment_service/src/main.rs:99:5
15: core::ops::function::FnOnce::call_once
at /rustc/82e1608dfa6e0b5569232559e3d385fea5a93112/library/core/src/ops/function.rs:250:5
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.
```
The attachment_service handles SIGQUIT by just exiting the process.
In theory, the SIGQUIT could come in while we're writing out the
`attachments.json`.
Now, in above log output, there's a 1 second gap between the last
request completing
and the SIGQUIT coming in. So, there must be some other issue.
But, let's have this change anyways, maybe it helps uncover the real
cause for the test failure.
1. Introduce a naive `Timeline::get_vectored` implementation
The return type is intended to be flexible enough for various types of
callers. We return the pages in a map keyed by `Key` such that the
caller doesn't have to map back to the key if it needs to know it. Some
callers can ignore errors
for specific pages, so we return a separate `Result<Bytes,
PageReconstructError>` for each page and an overarching
`GetVectoredError` for API misuse. The overhead of the mapping will be
small and bounded since we enforce a maximum key count for the
operation.
2. Use the `get_vectored` API for SLRU segment reconstruction and image
layer creation.
## Problem
Parsing the IP address at check time is a little wasteful.
## Summary of changes
Parse the IP when we get it from cplane. Adding a `None` variant to
still allow malformed patterns
## Problem
There are a lot of responses with 404 role not found error, which are
not getting cached in proxy.
## Summary of changes
If there was returned an empty secret but with the project_id, store it
in cache.
## Problem
There is "neon.pageserver_connstring" GUC with PGC_SIGHUP option,
allowing to change it using
pg_reload_conf(). It is used by control plane to update pageserver
connection string if page server is crashed,
relocated or new shards are added.
It is copied to shared memory because config can not be loaded during
query execution and we need to
reestablish connection to page server.
## Summary of changes
Copying connection string to shared memory is done by postmaster. And
other backends
should check update counter to determine of connection URL is changed
and connection needs to be reestablished.
We can not use standard Postgres LW-locks, because postmaster has proc
entry and so can not wait
on this primitive. This is why lockless access algorithm is implemented
using two atomic counters to enforce
consistent reading of connection string value from shared memory.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: Heikki Linnakangas <heikki@neon.tech>
The idea is to achieve separation between keyspace layout definition
and operating on said keyspace. I've inlined all these function since
they're small and we don't use LTO in the storage release builds
at the moment.
Closes https://github.com/neondatabase/neon/issues/6347
Before this patch, we would update the `tenant_state.intent` in memory
but not persist the detachment to disk.
I noticed this in https://github.com/neondatabase/neon/pull/6214 where
we stop, then restart, the attachment service.
## Problem
When a tenant is in Attaching state, and waiting for the
`concurrent_tenant_warmup` semaphore, it also listens for the tenant
cancellation token. When that token fires, Tenant::attach drops out.
Meanwhile, Tenant::set_stopping waits forever for the tenant to exit
Attaching state.
Fixes: https://github.com/neondatabase/neon/issues/6423
## Summary of changes
- In the absence of a valid state for the tenant, it is set to Broken in
this path. A more elegant solution will require more refactoring, beyond
this minimal fix.
## Problem
If you build the compute-node dockerfile with the PG_VERSION argument
passed in (e.g. `docker build -f Dockerfile.compute-node --build-arg
PG_VERSION=v15 .`, it fails, as some of stages doesn't have the
PG_VERSION arg defined.
## Summary of changes
Added the PG_VERSION arg to the plv8-build, neon-pg-ext-build, and
pg-embedding-pg-build stages of Dockerfile.compute-node
## Problem
See https://neondb.slack.com/archives/C06F5UJH601/p1705731304237889
Adding 1 to xid in `update_next_xid` can cause overflow in debug mode.
0xffffffff is valid transaction ID.
## Summary of changes
Use `wrapping_add`
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: Heikki Linnakangas <heikki@neon.tech>
## Problem
In https://github.com/neondatabase/neon/pull/6283 I did a couple changes
that weren't directly related to the goal of extracting the state
machine, so I'm putting them here
## Summary of changes
- move postgres vs console provider into another enum
- reduce error cases for link auth
- slightly refactor link flow
## Problem
Currently we store in cache even if the project is undefined. That makes
invalidation impossible.
## Summary of changes
Do not store if project id is empty.
update_next_xid() doesn't have any special treatment for the invalid or
other special XIDs, so it will treat InvalidTransactionId (0) as a
regular XID. If old nextXid is smaller than 2^31, 0 will look like a
very old XID, and nothing happens. But if nextXid is greater than 2^31 0
will look like a very new XID, and update_next_xid() will incorrectly
bump up nextXID.
configures nextest to kill tests after 1 minute. slow period is set to
20s which is how long our tests currently take in total, there will be 2
warnings and then the test will be killed and it's output logged.
Cc: #6361
Cc: #6368 -- likely this will be enough for longer time, but it will be
counter productive when we want to attach and debug; the added line
would have to be commented out.
## Problem
Some fields were missed in the initial spec.
## Summary of changes
Adds a success boolean (defaults to false unless specifically marked as
successful).
Adds a duration_us integer that tracks how many microseconds were taken
from session start through to request completion.
From #6037 on, until this patch, if the client opens the connection but
doesn't send a `PagestreamFeMessage` within the first 10ms, we'd close
the connection because `self.timeline_cancelled()` returns.
It returns because `self.shard_timelines` is still empty at that point:
it gets filled lazily within the handlers for the incoming messages.
Changes
-------
The question is: if we can't check for timeline cancellation, what else
do we need to be cancellable for? `tenant.cancel` is also a bad choice
because the `tenant` (shard) we pick at the top of handle_pagerequests
might indeed go away over the course of the connection lifetime, but
other shards may still be there.
The correct solution, I think, is to be responsive to task_mgr
cancellation, because the connection handler runs in a task_mgr task and
it is already the current canonical way how we shut down a tenant's /
timelin's page_service connections (see `Tenant::shutdown` /
`Timeline::shutdown`).
So, rename the function and make it sensitive to task_mgr cancellation.
In the most straightforward way; safekeeper performs it in DELETE endpoint
implementation, with no coordination between sks.
delete_force endpoint in the code is renamed to delete as there is only one way
to delete.
## Problem
For PRs with `run-benchmarks` label, we don't upload results to the db,
making it harder to debug such tests. The only way to see some
numbers is by examining GitHub Action output which is really
inconvenient.
This PR adds zenbenchmark metrics to Allure reports.
## Summary of changes
- Create a json file with zenbenchmark results and attach it to allure
report
In
7f828890cf
we changed the logic for persisting control_files. Previously it was
updated if `peer_horizon_lsn` jumped more than one segment, which made
`peer_horizon_lsn` initialized on disk as soon as safekeeper has
received a first `AppendRequest`.
This caused an issue with `truncateLsn`, which now can be zero
sometimes. This PR fixes it, and now `truncateLsn/peer_horizon_lsn` can
never be zero once we know `timeline_start_lsn`.
Closes https://github.com/neondatabase/neon/issues/6248
With testing the new eviction order there is a problem of all of the
(currently rare) disk usage based evictions being rare and unique; this
PR adds a human readable summary of what absolute order would had done
and what the relative order does. Assumption is that these loggings will
make the few evictions runs in staging more useful.
Cc: #5304 for allowing testing in the staging
## Problem
Use [NEON_SMGR] for all log messages produced by neon extension.
## Summary of changes
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
- Start pgbouncer in VM from postgres user, to allow connection to
pgbouncer admin console.
- Remove unused compute_ctl options --pgbouncer-connstr
and --pgbouncer-ini-path.
- Fix and cleanup code of connection to pgbouncer, add retries
because pgbouncer may not be instantly ready when compute_ctl starts.
## Problem
tenant_id/timeline_id is no longer a full identifier for metrics from a
`Tenant` or `Timeline` object.
Closes: https://github.com/neondatabase/neon/issues/5953
## Summary of changes
Include `shard_id` label everywhere we have `tenant_id`/`timeline_id`
label.
I just failed to see this earlier on #6136. layer counts are used as an
abstraction, and each of the two tenants lose proportionally about the
same amount of layers. sadly there is no difference in between
`relative_spare` and `relative_equal` as both of these end up evicting
the exact same amount of layers, but I'll try to add later another test
for those.
Cc: #5304
## Problem
Currently relation hash size is limited by "neon.relsize_hash_size" GUC
with default value 64k.
64k relations is not so small number... but it is enough to create 376
databases to exhaust it.
## Summary of changes
Use LRU replacement algorithm to prevent hash overflow
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
## Problem
To test sharding, we need something to control it. We could write python
code for doing this from the test runner, but this wouldn't be usable
with neon_local run directly, and when we want to write tests with large
number of shards/tenants, Rust is a better fit efficiently handling all
the required state.
This service enables automated tests to easily get a system with
sharding/HA without the test itself having to set this all up by hand:
existing tests can be run against sharded tenants just by setting a
shard count when creating the tenant.
## Summary of changes
Attachment service was previously a map of TenantId->TenantState, where
the principal state stored for each tenant was the generation and the
last attached pageserver. This enabled it to serve the re-attach and
validate requests that the pageserver requires.
In this PR, the scope of the service is extended substantially to do
overall management of tenants in the pageserver, including
tenant/timeline creation, live migration, evacuation of offline
pageservers etc. This is done using synchronous code to make declarative
changes to the tenant's intended state (`TenantState.policy` and
`TenantState.intent`), which are then translated into calls into the
pageserver by the `Reconciler`.
Top level summary of modules within
`control_plane/attachment_service/src`:
- `tenant_state`: structure that represents one tenant shard.
- `service`: implements the main high level such as tenant/timeline
creation, marking a node offline, etc.
- `scheduler`: for operations that need to pick a pageserver for a
tenant, construct a scheduler and call into it.
- `compute_hook`: receive notifications when a tenant shard is attached
somewhere new. Once we have locations for all the shards in a tenant,
emit an update to postgres configuration via the neon_local `LocalEnv`.
- `http`: HTTP stubs. These mostly map to methods on `Service`, but are
separated for readability and so that it'll be easier to adapt if/when
we switch to another RPC layer.
- `node`: structure that describes a pageserver node. The most important
attribute of a node is its availability: marking a node offline causes
tenant shards to reschedule away from it.
This PR is a precursor to implementing the full sharding service for
prod (#6342). What's the difference between this and a production-ready
controller for pageservers?
- JSON file persistence to be replaced with a database
- Limited observability.
- No concurrency limits. Marking a pageserver offline will try and
migrate every tenant to a new pageserver concurrently, even if there are
thousands.
- Very simple scheduler that only knows to pick the pageserver with
fewest tenants, and place secondary locations on a different pageserver
than attached locations: it does not try to place shards for the same
tenant on different pageservers. This matters little in tests, because
picking the least-used pageserver usually results in round-robin
placement.
- Scheduler state is rebuilt exhaustively for each operation that
requires a scheduler.
- Relies on neon_local mechanisms for updating postgres: in production
this would be something that flows through the real control plane.
---------
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
## Problem
The `/v1/tenant` listing API only applies to attached tenants.
For an external service to implement a global reconciliation of its list
of shards vs. what's on the pageserver, we need a full view of what's in
TenantManager, including secondary tenant locations, and InProgress
locations.
Dependency of https://github.com/neondatabase/neon/pull/6251
## Summary of changes
- Add methods to Tenant and SecondaryTenant to reconstruct the
LocationConf used to create them.
- Add `GET /v1/location_config` API
Previously, if we:
1. created a new timeline B from a different timeline's A initdb
2. deleted timeline A
the initdb for timeline B would be gone, at least in a world where we
are deleting initdbs upon timeline deletion. This world is imminent
(#6226).
Therefore, if the pageserver is instructed to load the initdb from a
different timeline ID, copy it to the newly created timeline's directory
in S3. This ensures that we can disaster recover the new timeline as
well, regardless of whether the original timeline was deleted or not.
Part of https://github.com/neondatabase/neon/issues/5282.
The remote_storage crate contains two copies of each test, one for azure
and one for S3. The repetition is not necessary and makes the tests more
prone to drift, so we remove it by moving the tests into a shared
module.
The module has a different name depending on where it is included, so
that each test still has "s3" or "azure" in its full path, allowing you
to just test the S3 test or just the azure tests.
Earlier PR that removed some duplication already: #6176Fixes#6146.
This implements the `copy` operation for azure blobs, added to S3 by
#6091, and adds tests both to s3 and azure ensuring that the copy
operation works.
Follows #6123
Closes: https://github.com/neondatabase/neon/issues/5342
The approach here is to avoid using `Layer` from secondary tenants, and
instead make the eviction types (e.g. `EvictionCandidate`) have a
variant that carries a Layer for attached tenants, and a different
variant for secondary tenants.
Other changes:
- EvictionCandidate no longer carries a `Timeline`: this was only used
for providing a witness reference to remote timeline client.
- The types for returning eviction candidates are all in
disk_usage_eviction_task.rs now, whereas some of them were in
timeline.rs before.
- The EvictionCandidate type replaces LocalLayerInfoForDiskUsageEviction
type, which was basically the same thing.
## Problem
In #5980 the page service connection handler gets a simple piece of
logic for finding the right Timeline: at connection time, it picks an
arbitrary Timeline, and then when handling individual page requests it
checks if the original timeline is the correct shard, and if not looks
one up.
This is pretty slow in the case where we have to go look up the other
timeline, because we take the big tenants manager lock.
## Summary of changes
- Add a `shard_timelines` map of ShardIndex to Timeline on the page
service connection handler
- When looking up a Timeline for a particular ShardIndex, consult
`shard_timelines` to avoid hitting the TenantsManager unless we really
need to.
- Re-work the CancellationToken handling, because the handler now holds
gateguards on multiple timelines, and so must respect cancellation of
_any_ timeline it has in its cache, not just the timeline related to the
request it is currently servicing.
---------
Co-authored-by: Vlad Lazar <vlad@neon.tech>
The theme of the changes in this PR is that they're enablers for #6251
which are superficial struct/api changes.
This is a spinoff from #6251:
- Various APIs + clients thereof take TenantShardId rather than TenantId
- The creation API gets a ShardParameters member, which may be used to
configure shard count and stripe size. This enables the attachment
service to present a "virtual pageserver" creation endpoint that creates
multiple shards.
- The attachment service will use tenant size information to drive shard
splitting. Make a version of `TenantHistorySize` that is usable for
decoding these API responses.
- ComputeSpec includes a shard stripe size.
## Problem
Se.e
https://github.com/orgs/neondatabase/projects/49/views/13?pane=issue&itemId=48282912
## Summary of changes
Do not suspend compute if there are active auto vacuum workers
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Fixes the race condition between timeline creation and tenant deletion
outlined in #6255.
Related: #5914, which is a similar race condition about the uninit
marker file.
Fixes#6255
Generally useful when debugging / troubleshooting.
I found this useful when manually duplicating a tenant from a script[^1]
where I can't use `neon_fixtures.Pageserver.tenant_attach`'s automatic
integration with the neon_local's attachment_service.
[^1]: https://github.com/neondatabase/neon/pull/6349
## Problem
Currently, activity monitor in `compute_ctl` has 500 ms polling
interval. It also looks on the list of current client backends looking
for an active one or one with the most recent state change. This means
we can miss short-living connections.
Yet, during testing this PR I realized that it's usually not a problem
with pooled connection, as pgbouncer maintains connections to Postgres
even though client connection are short-living. We can still miss direct
connections.
## Summary of changes
This commit introduces another way to detect user activity on the
compute. It polls a sum of `active_time` and sum of `sessions` from all
non-system databases in the `pg_stat_database` [1]. If user runs some
queries or just open a direct connection, it will rise; if user will
drop db, it can go down, but it's still a change and will be detected as
activity.
New statistic-based logic seems to be working fine. Yet, after having it
running for a couple of hours I've seen several odd cases with
connections via pgbouncer:
1. Sometimes, if you run just `psql pooler_connstr -c 'select 1;'`
`active_time` could be not updated immediately, and it may take a couple
of dozens of seconds. This doesn't seem critical, though.
2. Same query with pooler, `active_time` can be bumped a bit, then
pgbouncer keeps open connection to Postgres for ~10 minutes, then it
disconnects, and `active_time` *could be* bumped a bit again. 'Could be'
because I've seen it once, but it didn't reproduce for a second try.
I think this can create false-positives (hopefully rare), when we will
not suspend some computes because of lagged statistics update OR because
some non-user processes will try to connect to user databases.
Currently, we don't touch them outside of startup and
`postgres_exporter` is configured to do not discover other databases,
but this can change in the future.
New behavior is covered by feature flag `activity_monitor_experimental`,
which should be provided by control plane via neondatabase/cloud#9171
[1] https://www.postgresql.org/docs/current/monitoring-stats.html#MONITORING-PG-STAT-DATABASE-VIEW
Related to neondatabase/cloud#7966, neondatabase/cloud#7198
## Problem
When creating a timeline on a sharded tenant, we call into each shard.
We don't need to upload the initdb from every shard: only do it on shard
zero.
## Summary of changes
- Move the initdb upload into a function, and only call it on shard
zero.
## Problem
Occasional test failures with QueryError::Other errors saying
"cancelled" that get logged at error severity.
## Summary of changes
Avoid casting GetActiveTenantError::Cancelled into QueryError::Other --
it should be QueryError::Shutdown, which is not logged as an error.
Also shuts down `Broker`, which, before this PR, we did start in
`start()` but relied on the fixture to stop. Do it a bit earlier so
that, after `NeonEnv.stop()` returns, there are no child processes using
`repo_dir`.
Also, drive-by-fixes inverted logic around `ps_assert_metric_no_errors`,
missed during https://github.com/neondatabase/neon/pull/6295
---------
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
safekeeper.rs is mostly about consensus, but state is wider. Also form
SafekeeperState which encapsulates persistent part + in memory layer with API
for atomic updates.
Moves remote_consistent_lsn back to SafekeeperMemState, fixes its absense from
memory dump.
Also renames SafekeeperState to TimelinePersistentState, as TimelineMemState and
TimelinePersistent state are created.
## Problem
Previously, `GET /v1/tenant/:tenant_id/timeline` and `GET
/v1/tenant/:tenant_id/timeline/:timeline_id`
would bump the priority of the background task which computes the
initial logical size by cancelling
the wait on the synchronisation semaphore. However, the request would
still return an approximate
logical size. It's undesirable to force background work for a status
request.
## Summary of changes
This PR updates the priority used by the timeline status request such
that they don't do priority boosting
by default anymore. An optional query parameter,
`force-await-initial-logical-size`, is added for both
mentioned endpoints. When set to true, it will skip the concurrency
limiting semaphore and wait
for the background task to complete before returning the exact logical
size.
In order to exercise this behaviour in a test I had to add an extra
failpoint. If you think it's too intrusive,
it can be removed.
Also fixeda small bug where the cancellation of a download is reported as an
opaque download failure upstream. This caused `test_location_conf_churn`
to fail at teardown due to a WARN log line.
Closes https://github.com/neondatabase/neon/issues/6168
## Problem
HTTP connection pool was not respecting the PitR options.
## Summary of changes
1. refactor neon_options a bit to allow easier access to cache_key
2. make HTTP not go through `StartupMessageParams`
3. expose SNI processing to replace what was removed in step 2.
Random find while looking at an idle 20k tenant pageserver where each
tenant
has 9 tiny L0 layers and compaction produces no new L1s / image layers.
The aggregate CPU cost of running this every 20s for 20k tenants is
actually substantial, due to the use of `spawn_blocking`.
This reverts commit ab1f37e908.
Thereby
fixes#5479
Updated Analysis
================
The problem with the original patch was that it, for the first time,
exposed the `VirtualFile` code to tokio task concurrency instead of just
thread-based concurrency. That caused the VirtualFile file descriptor
cache to start thrashing, effectively grinding the system to a halt.
Details
-------
At the time of the original patch, we had a _lot_ of runnable tasks in
the pageserver.
The symptom that prompted the revert (now being reverted in this PR) is
that our production systems fell into a valley of zero goodput, high
CPU, and zero disk IOPS shortly after PS restart.
We lay out the root cause for that behavior in this subsection.
At the time, there was no concurrency limit on the number of concurrent
initial logical size calculations.
Initial size calculation was initiated for all timelines within the
first 10 minutes as part of consumption metrics collection.
On a PS with 20k timelines, we'd thus have 20k runnable tasks.
Before the original patch, the `VirtualFile` code never returned
`Poll::Pending`.
That meant that once we entered it, the calling tokio task would not
yield to the tokio executor until we were done performing the
VirtualFile operation, i.e., doing a blocking IO system call.
The original patch switched the VirtualFile file descriptor cache's
synchronization primitives to those from `tokio::sync`.
It did not change that we were doing synchronous IO system calls.
And the cache had more slots than we have tokio executor threads.
So, these primitives never actually needed to return `Poll::Pending`.
But, the tokio scheduler makes tokio sync primitives return `Pending`
*artificially*, as a mechanism for the scheduler to get back into
control more often
([example](https://docs.rs/tokio/1.35.1/src/tokio/sync/batch_semaphore.rs.html#570)).
So, the new reality was that VirtualFile calls could now yield to the
tokio executor.
Tokio would pick one of the other 19999 runnable tasks to run.
These tasks were also using VirtualFile.
So, we now had a lot more concurrency in that area of the code.
The problem with more concurrency was that caches started thrashing,
most notably the VirtualFile file descriptor cache: each time a task
would be rescheduled, it would want to do its next VirtualFile
operation. For that, it would first need to evict another (task's)
VirtualFile fd from the cache to make room for its own fd. It would then
do one VirtualFile operation before hitting an await point and yielding
to the executor again. The executor would run the other 19999 tasks for
fairness before circling back to the first task, which would find its fd
evicted.
The other cache that would theoretically be impacted in a similar way is
the pageserver's `PageCache`.
However, for initial logical size calculation, it seems much less
relevant in experiments, likely because of the random access nature of
initial logical size calculation.
Fixes
=====
We fixed the above problems by
- raising VirtualFile cache sizes
- https://github.com/neondatabase/cloud/issues/8351
- changing code to ensure forward-progress once cache slots have been
acquired
- https://github.com/neondatabase/neon/pull/5480
- https://github.com/neondatabase/neon/pull/5482
- tbd: https://github.com/neondatabase/neon/issues/6065
- reducing the amount of runnable tokio tasks
- https://github.com/neondatabase/neon/pull/5578
- https://github.com/neondatabase/neon/pull/6000
- fix bugs that caused unnecessary concurrency induced by connection
handlers
- https://github.com/neondatabase/neon/issues/5993
I manually verified that this PR doesn't negatively affect startup
performance as follows:
create a pageserver in production configuration, with 20k
tenants/timelines, 9 tiny L0 layer files each; Start it, and observe
```
INFO Startup complete (368.009s since start) elapsed_ms=368009
```
I further verified in that same setup that, when using `pagebench`'s
getpage benchmark at as-fast-as-possible request rate against 5k of the
20k tenants, the achieved throughput is identical. The VirtualFile cache
isn't thrashing in that case.
Future Work
===========
We will still exposed to the cache thrashing risk from outside factors,
e.g., request concurrency is unbounded, and initial size calculation
skips the concurrency limiter when we establish a walreceiver
connection.
Once we start thrashing, we will degrade non-gracefully, i.e., encounter
a valley as was seen with the original patch.
However, we have sufficient means to deal with that unlikely situation:
1. we have dashboards & metrics to monitor & alert on cache thrashing
2. we can react by scaling the bottleneck resources (cache size) or by
manually shedding load through tenant relocation
Potential systematic solutions are future work:
* global concurrency limiting
* per-tenant rate limiting => #5899
* pageserver-initiated load shedding
Related Issues
==============
This PR unblocks the introduction of tokio-epoll-uring for asynchronous
disk IO ([Epic](#4744)).
Postgres can write multiline logs, and they are difficult to handle
after they are mixed with other logs. This PR combines multiline logs
from postgres into a single line, where previous line breaks are
replaced with unicode zero-width spaces. Then postgres logs are written
to stderr with `PG:` prefix.
It makes it easy to distinguish postgres logs from all other compute
logs with a simple grep, e.g. `|= "PG:"`
## Problem
Current cache doesn't support any updates from the cplane.
## Summary of changes
* Added redis notifier listner.
* Added cache which can be invalidated with the notifier. If the
notifier is not available, it's just a normal ttl cache.
* Updated cplane api.
The motivation behind this organization of the data is the following:
* In the Neon data model there are projects. Projects could have
multiple branches and each branch could have more than one endpoint.
* Also there is one special `main` branch.
* Password reset works per branch.
* Allowed IPs are the same for every branch in the project (except,
maybe, the main one).
* The main branch can be changed to the other branch.
* The endpoint can be moved between branches.
Every event described above requires some special processing on the
porxy (or cplane) side.
The idea of invalidating for the project is that whenever one of the
events above is happening with the project, proxy can invalidate all
entries for the entire project.
This approach also requires some additional API change (returning
project_id inside the auth info).
PR #6266 broke the getpage_latest_lsn benchmark.
Before this patch, we'd fail with
```
not implemented: split up range
```
because `r.start = rel size key` and `r.end = rel size key + 1`.
The filtering of the key ranges in that loop is a bit ugly, but,
I measured:
* setup with 180k layer files (20k tenants * 9 layers).
* total physical size is 463GiB
* 5k tenants, the range filtering takes `0.6 seconds` on an
i3en.3xlarge.
That's a tiny fraction of the overall time it takes for pagebench to get
ready to send requests. So, this is good enough for now / there are
other bottlenecks that are bigger.
## Problem
The code for tenant create and tenant attach was just a special case of
what upsert_location does.
## Summary of changes
- Use `upsert_location` for create and attach APIs
- Clean up error handling in upsert_location so that it can generate
appropriate HTTP response codes
- Update tests that asserted the old non-idempotent behavior of attach
- Rework the `test_ignore_while_attaching` test, and fix tenant shutdown
during activation, which this test was supposed to cover, but it was
actually just waiting for activation to complete.
This uses the [newly stable](https://blog.rust-lang.org/2023/12/21/async-fn-rpit-in-traits.html)
async trait feature for three internal traits. One requires `Send`
bounds to be present so uses `impl Future<...> + Send` instead.
Advantages:
* less macro usage
* no extra boxing
Disadvantages:
* impl syntax needed for `Send` bounds is a bit more verbose (but only
required in one place)
Part of #5771
Extracted from https://github.com/neondatabase/neon/pull/6214
This PR makes the test suite sensitive to the new env var
`NEON_ENV_BUILDER_FROM_REPO_DIR_USE_OVERLAYFS`.
If it is set, `NeonEnvBuilder.from_repo_dir` uses overlayfs
to duplicate the the snapshot repo dir contents.
Since mounting requires root privileges, we use sudo to perform
the mounts. That, and macOS support, is also why copytree remains
the default.
If we ever run on a filesystem with copy reflink support, we should
consider that as an alternative.
This PR can be tried on a Linux machine on the
`test_backward_compatiblity` test, which uses `from_repo_dir`.
During a previous incident, we noticed that this particular line can be
repeatedly logged every 100ms if the memory usage continues is
persistently high enough to warrant upscaling.
Per the added comment: Ideally we'd still like to include this log line,
because it's useful information, but the simple way to include it
produces far too many log lines, and the more complex ways to
deduplicate the log lines while still including the information are
probably not worth the effort right now.
This fixes the clippy lint firing on macOS on the conversion which
needed for portability. For some reason, the logic in
https://github.com/rust-lang/rust-clippy/pull/11669 to avoid an overlap
is not working.
## Summary of changes
### RequestMonitoring
We want to add an event stream with information on each request for
easier analysis than what we can do with diagnostic logs alone
(https://github.com/neondatabase/cloud/issues/8807). This
RequestMonitoring will keep a record of the final state of a request. On
drop it will be pushed into a queue to be uploaded.
Because this context is a bag of data, I don't want this information to
impact logic of request handling. I personally think that weakly typed
data (such as all these options) makes for spaghetti code. I will
however allow for this data to impact rate-limiting and blocking of
requests, as this does not _really_ change how a request is handled.
### Parquet
Each `RequestMonitoring` is flushed into a channel where it is converted
into `RequestData`, which is accumulated into parquet files. Each file
will have a certain number of rows per row group, and several row groups
will eventually fill up the file, which we then upload to S3.
We will also upload smaller files if they take too long to construct.
This is a lightweight change to keep the scrubber providing sensible
output when using sharding.
- The timeline count was wrong when using sharding
- When checking for tenant existence, we didn't re-use results between
different shards in the same tenant
Closes: https://github.com/neondatabase/neon/issues/5929
These functions don't need updating for sharding: it's fine for them to
remain shard-naive, as they're only used in the context of dumping a
layer file. The sharding metadata doesn't live in the layer file, it
lives in the index.
## Problem
Noticed while making other changes that there were `pub` items that were
unused.
## Summary of changes
- Make everything `pub(crate)` in metrics.rs, apart from items used from
`bin/`
- Fix the timelines eviction metric: it was never being incremented
- Remove an unused ephemeral_bytes counter.
## Problem
`black` is slow sometimes, we can replace it with `ruff format` (a new
feature in 0.1.2 [0]), which produces pretty similar to black style [1].
On my local machine (MacBook M1 Pro 16GB):
```
# `black` on main
$ hyperfine "BLACK_CACHE_DIR=/dev/null poetry run black ."
Benchmark 1: BLACK_CACHE_DIR=/dev/null poetry run black .
Time (mean ± σ): 3.131 s ± 0.090 s [User: 5.194 s, System: 0.859 s]
Range (min … max): 3.047 s … 3.354 s 10 runs
```
```
# `ruff format` on the current PR
$ hyperfine "RUFF_NO_CACHE=true poetry run ruff format"
Benchmark 1: RUFF_NO_CACHE=true poetry run ruff format
Time (mean ± σ): 300.7 ms ± 50.2 ms [User: 259.5 ms, System: 76.1 ms]
Range (min … max): 267.5 ms … 420.2 ms 10 runs
```
## Summary of changes
- Replace `black` with `ruff format` everywhere
- [0] https://docs.astral.sh/ruff/formatter/
- [1] https://docs.astral.sh/ruff/formatter/#black-compatibility
Before this PR, `is_rel_block_key` returns true for the blknum
`0xffffffff`,
which is a blknum that's actually never written by Postgres, but used by
Neon Pageserver to store the relsize.
Quoting @MMeent:
> PostgreSQL can't extend the relation beyond size of 0xFFFFFFFF blocks,
> so block number 0xFFFFFFFE is the last valid block number.
This PR changes the definition of the function to exclude blknum
0xffffffff.
My motivation for doing this change is to fix the `pagebench` getpage
benchmark, which uses `is_rel_block_key` to filter the keyspace for
valid pages to request from page_service.
fixes https://github.com/neondatabase/neon/issues/6210
I checked other users of the function.
The first one is `key_is_shard0`, which already had added an exemption
for 0xffffffff. So, there's no functional change with this PR.
The second one is `DatadirModification::flush`[^1]. With this PR,
`.flush()` will skip the relsize key, whereas it didn't
before. This means we will pile up all the relsize key-value pairs
`(Key,u32)`
in `DatadirModification::pending_updates` until `.commit()` is called.
The only place I can think of where that would be a problem is if we
import from a full basebackup, and don't `.commit()` regularly,
like we currently don't do in `import_basebackup_from_tar`.
It exposes us to input-controlled allocations.
However, that was already the case for the other keys that are skipped,
so, one can argue that this change is not making the situation much
worse.
[^1]: That type's `flush()` and `commit()` methods are terribly named,
but,
that's for another time
Implement API for cloning a single timeline inside a safekeeper. Also
add API for calculating a sha256 hash of WAL, which is used in tests.
`/copy` API works by copying objects inside S3 for all but the last
segments, and the last segments are copied on-disk. A special temporary
directory is created for a timeline, because copy can take a lot of
time, especially for large timelines. After all files segments have been
prepared, this directory is mounted to the main tree and timeline is
loaded to memory.
Some caveats:
- large timelines can take a lot of time to copy, because we need to
copy many S3 segments
- caller should wait for HTTP call to finish indefinetely and don't
close the HTTP connection, because it will stop the process, which is
not continued in the background
- `until_lsn` must be a valid LSN, otherwise bad things can happen
- API will return 200 if specified `timeline_id` already exists, even if
it's not a copy
- each safekeeper will try to copy S3 segments, so it's better to not
call this API in-parallel on different safekeepers
## Problem
- When a client requests a key that isn't found in any shard on the node
(edge case that only happens if a compute's config is out of date), we
should prompt them to reconnect (as this includes a backoff), since they
will not be able to complete the request until they eventually get a
correct pageserver connection string.
- QueryError::Other is used excessively: this contains a type-ambiguous
anyhow::Error and is logged very verbosely (including backtrace).
## Summary of changes
- Introduce PageStreamError to replace use of anyhow::Error in request
handlers for getpage, etc.
- Introduce Reconnect and NotFound variants to QueryError
- Map the "shard routing error" case to PageStreamError::Reconnect ->
QueryError::Reconnect
- Update type conversions for LSN timeouts and tenant/timeline not found
errors to use PageStreamError::NotFound->QueryError::NotFound
Previously, we would wait for the LSN to be visible on whichever
timeline we happened to load at the start of the connection, then
proceed to look up the correct timeline for the key and do the read.
If the timeline holding the key was behind the timeline we used
for the LSN wait, then we might serve an apparently-successful read result
that actually contains data from behind the requested lsn.
If there is some secondary shard for a tenant on the same
node as an attached shard, the secondary shard could trip up
this code and cause page_service to incorrectly
get an error instead of finding the attached shard.
It has caveats such as creating half empty segment which can't be
offloaded. Instead we'll pursue approach of pull_timeline, seeding new state
from some peer.
## Problem
We need to add one more patch to pgbouncer (for
https://github.com/neondatabase/neon/issues/5801). I've decided to
cherry-pick all required patches to a pgbouncer fork
(`neondatabase/pgbouncer`) and use it instead.
See
https://github.com/neondatabase/pgbouncer/releases/tag/pgbouncer_1_21_0-neon-1
## Summary of changes
- Revert the previous patch (for deallocate/discard all) — the fork
already contains it.
- Remove `libssl-dev` dependency — we build pgbouncer without `openssl`
support.
- Clone git tag and build pgbouncer from source code.
## Problem
For context, this problem was observed in a research project where we
try to make neon run in multiple regions and I was asked by @hlinnaka to
make this PR.
In our project, we use the pageserver in a non-conventional way such
that we would send a larger number of requests to the pageserver than
normal (imagine postgres without the buffer pool). I measured the time
from the moment a WAL record left the safekeeper to when it reached the
pageserver
([code](e593db1f5a/pageserver/src/tenant/timeline/walreceiver/walreceiver_connection.rs (L282-L287)))
and observed that when the number of get_page_at_lsn requests was high,
the wal receiving time increased significantly (see the left side of the
graphs below).
Upon further investigation, I found that the delay was caused by this
line
d2ca410919/pageserver/src/tenant/timeline.rs (L2348)
The `get_layer_for_write` method is called for every value during WAL
ingestion and it tries to acquire layers write lock every time, thus
this results in high contention when read lock is acquired more
frequently.
## Summary of changes
It is unnecessary to call `get_layer_for_write` repeatedly for all
values in a WAL message since they would end up in the same memory layer
anyway, so I created the batched versions of `InMemoryLayer::put_value`,
`InMemoryLayer ::put_tombstone`, `Timeline::put_value`, and
`Timeline::put_tombstone`, that acquire the locks once for a batch of
values.
Additionally, `DatadirModification` is changed to store multiple
versions of uncommitted values, and `WalIngest::ingest_record()` can now
ingest records without immediately committing them.
With these new APIs, the new ingestion loop can be changed to commit for
every `ingest_batch_size` records. The `ingest_batch_size` variable is
exposed as a config. If it is set to 1 then we get the same behavior
before this change. I found that setting this value to 100 seems to work
the best, and you can see its effect on the right side of the above
graphs.
---------
Co-authored-by: John Spray <john@neon.tech>
## Problem
The version of pytest we were using emits a number of
DeprecationWarnings on latest python: these are fixed in latest release.
boto3 and python-dateutil also have deprecation warnings, but
unfortunately these aren't fixed upstream yet.
## Summary of changes
- Update pytest
- Update boto3 (this doesn't fix deprecation warnings, but by the time I
figured that out I had already done the update, and it's good hygiene
anyway)
The tool still needs a lot of work. These are the easiest fix and
feature:
- use similar adaptive config with s3 as remote_storage, use retries
- process only particular tenants
Tenants need to be from the correct region, they are not deduplicated,
but the feature is useful for re-checking small amount of tenants after
a large run.
To exercise MAX_SEND_SIZE sending from safekeeper; we've had a bug with WAL
records torn across several XLogData messages. Add failpoint to safekeeper to
slow down sending. Also check for corrupted WAL complains in standby log.
Make the test a bit simpler in passing, e.g. we don't need explicit commits as
autocommit is enabled by default.
https://neondb.slack.com/archives/C05L7D1JAUS/p1703774799114719https://github.com/neondatabase/cloud/issues/9057
Otherwise they are left orphaned when compute_ctl is terminated with a
signal. It was invisible most of the time because normally neon_local or k8s
kills postgres directly and then compute_ctl finishes gracefully. However, in
some tests compute_ctl gets stuck waiting for sync-safekeepers which
intentionally never ends because safekeepers are offline, and we want to stop
compute_ctl without leaving orphanes behind.
This is a quite rough approach which doesn't wait for children termination. A
better way would be to convert compute_ctl to async which would make waiting
easy.
## Problem
`cargo test` doesn't support timeouts
or junit output format
## Summary of changes
- Add `nextest` to `build-tools` image
- Switch `cargo test` with `cargo nextest` on CI
- Set timeout
## Problem
control plane should be able to return domain names and not just IP
addresses.
## Summary of changes
1. add regression tests
2. use rsplit to split the port from the back, then trim the ipv6
brackets
## Problem
Extension pg_stat_statements has function pg_stat_statements_reset().
In vanilla Postgres this function can only be called by superuser role
or other users/roles explicitly granted.
In Neon no end user can use superuser role.
Instead we have neon_superuser role.
We need to grant execute on pg_stat_statements_reset() to neon_superuser
## Summary of changes
Modify the Postgres v14, v15, v16 contrib in our compute docker file to
grant execute on pg_stat_statements_reset() to neon_superuser.
(Modifying it in our docker file is preferable to changes in
neondatabase/postgres because we want to limit the changes in our fork
that we have to carry with each new version of Postgres).
Note that the interface of proc/function pg_stat_statements_reset
changed in pg_stat_statements version 1.7
So for versions up to and including 1.6 we must
`GRANT EXECUTE ON FUNCTION pg_stat_statements_reset() TO
neon_superuser;`
and for versions starting from 1.7 we must
`GRANT EXECUTE ON FUNCTION pg_stat_statements_reset(Oid, Oid, bigint) TO
neon_superuser;`
If we just use `GRANT EXECUTE ON FUNCTION pg_stat_statements_reset() TO
neon_superuser;` for all version this results in the following error for
versions 1.7+:
```sql
neondb=> create extension pg_stat_statements;
ERROR: function pg_stat_statements_reset() does not exist
```
## Checklist before requesting a review
- [x ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [x ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
## I have run the following test and could now invoke
pg_stat_statements_reset() using default user
```bash
(neon) peterbendel@Peters-MBP neon % kubectl get pods | grep compute-quiet-mud-88416983
compute-quiet-mud-88416983-74f4bf67db-crl4c 3/3 Running 0 7m26s
(neon) peterbendel@Peters-MBP neon % kubectl set image deploy/compute-quiet-mud-88416983 compute-node=neondatabase/compute-node-v15:7307610371
deployment.apps/compute-quiet-mud-88416983 image updated
(neon) peterbendel@Peters-MBP neon % psql postgresql://peterbendel:<secret>@ep-bitter-sunset-73589702.us-east-2.aws.neon.build/neondb
psql (16.1, server 15.5)
SSL connection (protocol: TLSv1.3, cipher: TLS_AES_256_GCM_SHA384, compression: off)
Type "help" for help.
neondb=> select version();
version
---------------------------------------------------------------------------------------------------
PostgreSQL 15.5 on x86_64-pc-linux-gnu, compiled by gcc (Debian 10.2.1-6) 10.2.1 20210110, 64-bit
(1 row)
neondb=> create extension pg_stat_statements;
CREATE EXTENSION
neondb=> select pg_stat_statements_reset();
pg_stat_statements_reset
--------------------------
(1 row)
```
- add pgbouncer_settings section to compute spec;
- add pgbouncer-connstr option to compute_ctl.
- add pgbouncer-ini-path option to compute_ctl. Default: /etc/pgbouncer/pgbouncer.ini
Apply pgbouncer config on compute start and respec to override default spec.
Save pgbouncer config updates to pgbouncer.ini to preserve them across pgbouncer restarts.
Remove confirm_wal_streamed; we already apply both write and flush positions of
the slot to commit_lsn which is fine because 1) we need to wake up waiters 2)
committed WAL can be fetched from safekeepers by neon_walreader now.
wp -> sk communication now uses neon_walreader which will fetch missing WAL on
demand from safekeepers, so doesn't need this anymore. Also, cap WAL download by
max_slot_wal_keep_size to be able to start compute if lag is too high.
It is similar to XLogReader, but when either requested segment is missing
locally or requested LSN is before basebackup_lsn NeonWALReader asynchronously
fetches WAL from one of safekeepers.
Patch includes walproposer switch to NeonWALReader, splitting wouldn't make much
sense as it is hard to test otherwise. This finally removes risk of pg_wal
explosion (as well as slow start time) when one safekeeper is lagging, at the
same time allowing to recover it.
In the future reader should also be used by logical walsender for similar
reasons (currently we download the tail on compute start synchronously).
The main test is test_lagging_sk. However, I also run it manually a lot varying
MAX_SEND_SIZE on both sides (on safekeeper and on walproposer), testing various
fragmentations (one side having small buffer, another, both), which brought up
https://github.com/neondatabase/neon/issues/6055
closes https://github.com/neondatabase/neon/issues/1012
## Problem
See https://neondb.slack.com/archives/C026T7K2YP9/p1702813041997959
## Summary of changes
Do not take in account invalidated slots when calculate restart_lsn
position for basebackup at page server
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Store the content of the `last-modified` and `etag` HTTP headers in
`Download`.
This serves both as the first step towards #6199 and as a preparation
for tests in #6155 .
## Problem
During startup, a tenant delete request might have to retry for many
minutes waiting for a tenant to enter Active state.
## Summary of changes
- Refactor delete_tenant into TenantManager: this is not a functional
change, but will avoid merge conflicts with
https://github.com/neondatabase/neon/pull/6105 later
- Add 412 responses to the swagger definition of this endpoint.
- Use Tenant::wait_to_become_active in `TenantManager::delete_tenant`
---------
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
## Problem
Fix build-tools image tag calculation for PRs.
Broken in https://github.com/neondatabase/neon/pull/6195
## Summary of changes
- Use `pinned` tag instead of `$GITHUB_RUN_ID` if there's no changes in
the dockerfile (and we don't build such image)
## Fixing GitHub workflow issue related to build and push images
## Summary of changes
Followup of PR#608[move docker file from build repo to neon to solve
issue some issues
The build started failing because it missed a validation in logic that
determines changes in the docker file
Also, all the dependent jobs were skipped because of the build and push
of the image job.
To address the above issue following changes were made
- we are adding validation to generate image tag even if it's a merge to
repo.
- All the dependent jobs won't skip even if the build and push image job
is skipped.
- We have moved the logic to generate a tag in the sub-workflow. As the
tag name was necessary to be passed to the sub-workflow it made sense to
abstract that away where it was needed and then store it as an output
variable so that downward dependent jobs could access the value.
- This made the dependency logic easy and we don't need complex
expressions to check the condition on which it will run
- An earlier PR was closed that tried solving a similar problem that has
some feedback and context before creating this PR
https://github.com/neondatabase/neon/pull/6175
## Checklist before requesting a review
- [x] Move the tag generation logic from the main workflow to the
sub-workflow of build and push the image
- [x] Add a condition to generate an image tag for a non-PR-related run
- [x] remove complex if the condition from the job if conditions
---------
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
Co-authored-by: Abhijeet Patil <abhijeet@neon.tech>
## Problem
This is a comment only change.
To ensure that our benchmarking results are fair we need to have correct
stats in catalog. Otherwise optimizer chooses seq scan instead of index
only scan for some queries. Added comment to run vacuum after data prep.
This PR adds a component-level benchmarking utility for pageserver.
Its name is `pagebench`.
The problem solved by `pagebench` is that we want to put Pageserver
under high load.
This isn't easily achieved with `pgbench` because it needs to go through
a compute, which has signficant performance overhead compared to
accessing Pageserver directly.
Further, compute has its own performance optimizations (most
importantly: caches). Instead of designing a compute-facing workload
that defeats those internal optimizations, `pagebench` simply bypasses
them by accessing pageserver directly.
Supported benchmarks:
* getpage@latest_lsn
* basebackup
* triggering logical size calculation
This code has no automated users yet.
A performance regression test for getpage@latest_lsn will be added in a
later PR.
part of https://github.com/neondatabase/neon/issues/5771
## Problem
HTTP connection pool will grow without being pruned
## Summary of changes
Remove connection clients from pools once idle, or once they exit.
Periodically clear pool shards.
GC Logic:
Each shard contains a hashmap of `Arc<EndpointPool>`s.
Each connection stores a `Weak<EndpointPool>`.
During a GC sweep, we take a random shard write lock, and check that if
any of the `Arc<EndpointPool>`s are unique (using `Arc::get_mut`).
- If they are unique, then we check that the endpoint-pool is empty, and
sweep if it is.
- If they are not unique, then the endpoint-pool is in active use and we
don't sweep.
- Idle connections will self-clear from the endpoint-pool after 5
minutes.
Technically, the uniqueness of the endpoint-pool should be enough to
consider it empty, but the connection count check is done for
completeness sake.
Allows usage via `aws sso login --profile=<p>; AWS_PROFILE=<p>`. Now
there is no need to manually configure things any more via
`SSO_ACCOUNT_ID` and others. Now one can run the tests locally (given
Neon employee access to aws):
```
aws sso login --profile dev
export ENABLE_REAL_S3_REMOTE_STORAGE=nonempty REMOTE_STORAGE_S3_REGION=eu-central-1 REMOTE_STORAGE_S3_BUCKET=neon-github-public-dev AWS_PROFILE=dev
cargo test -p remote_storage -j 1 s3 -- --nocapture
```
Also makes the scrubber use the same region for auth that it does its
operations in (not touching the hard coded role name and start_url
values here, they are not ideal though).
Adds a new disk usage based eviction option, EvictionOrder, which
selects whether to use the current `AbsoluteAccessed` or this new
proposed but not yet tested `RelativeAccessed`. Additionally a fudge
factor was noticed while implementing this, which might help sparing
smaller tenants at the expense of targeting larger tenants.
Cc: #5304
Co-authored-by: Arpad Müller <arpad@neon.tech>
## Problem
Currently a chart of S3 error rate is misleading: it can show errors any
time we are attaching a tenant (probing for index_part generation,
checking for remote delete marker).
Considering 404 successful isn't perfectly elegant, but it enables the
error rate to be used a a more meaningful alert signal: it would
indicate if we were having auth issues, sending bad requests, getting
throttled ,etc.
## Summary of changes
Track 404 requests in the AttemptOutcome::Ok bucket instead of the
AttemptOutcome::Err bucket.
This is a precursor to:
- https://github.com/neondatabase/neon/pull/6185
While that PR contains big changes to neon_local and attachment_service,
this PR contains a few unrelated standalone changes generated while
working on that branch:
- Fix restarting a pageserver when it contains multiple shards for the
same tenant
- When using location_config api to attach a tenant, create its
timelines dir
- Update test paths where generations were previously optional to make
them always-on: this avoids tests having to spuriously assert that
attachment_service is not None in order to make the linter happy.
- Add a TenantShardId python implementation for subsequent use in test
helpers that will be made shard-aware
- Teach scrubber to read across shards when checking for layer
existence: this is a refactor to track the list of existent layers at
tenant-level rather than locally to each timeline. This is a precursor
to testing shard splitting.
Old methods are unchanged for backwards compatibility. Added
`SafekeeperDiscoveryRequest` and `SafekeeperDiscoveryResponse` types to
serve as example, and also as a prerequisite for
https://github.com/neondatabase/neon/issues/5471
It wasn't being incremented.
Fixup of
commit 1c88824ed0
Author: Christian Schwarz <christian@neon.tech>
Date: Fri Dec 1 12:52:59 2023 +0100
initial logical size calculation: add a bunch of metrics (#5995)
There is still default 'postgres' database, that may contain objects
owned by the role or some ACLs. We need to reassign objects in this
database too.
## Problem
If customer deleted all databases and then tries to delete role, that
has some non-standard ACLs,
`apply_config` operation will stuck because of failing role deletion.
## Problem
GCCproduce warning that bread result is not checked. It doesn't affect
program logic, but better live without warnings.
## Summary of changes
Check read result.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
## Problem
To understand differences in performance between neon, aurora and rds we
want to collect explain analyze plans and pg_stat_statements for
selected benchmarking runs
## Summary of changes
Add workflow input options to collect explain and pg_stat_statements for
benchmarking workflow
Co-authored-by: BodoBolero <bodobolero@gmail.com>
This squashes a particularly noisy warn-level log that occurs when
safekeepers are restarted.
Unfortunately the error type from `tonic` doesn't provide a neat way of
matching this, so we use a string comparison
Part of getpage@lsn benchmark epic:
https://github.com/neondatabase/neon/issues/5771
This allows getting the list of tenants and timelines without triggering
initial logical size calculation by requesting the timeline details API
response, which would skew our results.
The PR does two things:
* move the util functions present in the remote_storage Azure and S3
test files into a shared one, deduplicating them.
* add a `s3_upload_download_works` test as a copy of the Azure test
The goal is mainly to fight duplication and make the code a little bit
more generic (like removing mentions of s3 and azure from function
names).
This is a first step towards #6146.
## Problem
If the user reset password, cache could receive this information only
after `ttl` minutes.
## Summary of changes
Invalidate password on auth failure.
- No need to include c.h, port.h or pg_config.h, they are included in
postgres.h
- No need to include postgres.h in header files. Instead, the
assumption in PostgreSQL is that all .c files include postgres.h.
- Reorder includes to alphabetical order, and system headers before
pgsql headers
- Remove bunch of other unnecessary includes that got copy-pasted from
one source file to another
## Problem
Existing dependencies didn't work on Fedora 39 (python 3.12)
## Summary of changes
- Update pyyaml 6.0 -> 6.0.1
- Update yarl 1.8.2->1.9.4
- Update the `dnf install` line in README to include dependencies of
python packages (unrelated to upgrades, just noticed absences while
doing fresh pysync run)
## Problem
Currently if we are getting many consecutive connections to the same
user/ep we will send a lot of traffic to the console.
## Summary of changes
Cache with ttl=4min proxy_get_role_secret response.
Note: this is the temporary hack, notifier listener is WIP.
## Problem
The `src/proxy.rs` file is far too large
## Summary of changes
Creates 3 new files:
```
src/metrics.rs
src/proxy/retry.rs
src/proxy/connect_compute.rs
```
## Problem
#6112 added some logs and metrics: clean these up a bit:
- Avoid counting startup completions for tenants launched after startup
- exclude no-op cases from timing histograms
- remove a rogue log messages
Walproposer sometimes intentionally PANICs when its term is defeated as the
basebackup is likely spoiled by that time. We don't want core dumped in this
case.
It turns out the issue with skipped jobs is not so trivial (because
Github checks jobs transitively), a possible workaround with `if:
always() && contains(fromJSON('["success", "skipped"]'),
needs.build-buildtools-image.result)` will tangle the workflow really
bad. We'll need to come up with a better solution.
To unblock the main I'm going to revert
https://github.com/neondatabase/neon/pull/6082.
## Currently our build docker file is located in the build repo it makes
sense to have it as a part of our neon repo
## Summary of changes
We had the docker file that we use to build our binary and other tools
resided in the build repo
It made sense to bring the docker file to its repo where it has been
used
So that the contributors can also view it and amend if required
It will reduce the maintenance. Docker file changes and code changes can
be accommodated in same PR
Also, building the image and pushing it to ECR is abstracted in a
reusable workflow. Ideal is to use that for any other jobs too
## Checklist before requesting a review
- [x] Moved the docker file used to build the binary from the build repo
to the neon repo
- [x] adding gh workflow to build and push the image
- [x] adding gh workflow to tag the pushed image
- [x] update readMe file
---------
Co-authored-by: Abhijeet Patil <abhijeet@neon.tech>
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
## Problem
https://github.com/neondatabase/neon/security/dependabot/48
```
$ cargo tree -i zerocopy
zerocopy v0.7.3
└── ahash v0.8.5
└── hashbrown v0.13.2
```
ahash doesn't use the affected APIs we we are not vulnerable but best to
update to silence the alert anyway
## Summary of changes
```
$ cargo update -p zerocopy --precise 0.7.31
Updating crates.io index
Updating syn v2.0.28 -> v2.0.32
Updating zerocopy v0.7.3 -> v0.7.31
Updating zerocopy-derive v0.7.3 -> v0.7.31
```
## Problem
During startup, a client request might have to wait a long time while
the system is busy initializing all the attached tenants, even though
most of the attached tenants probably don't have any client requests to
service, and could wait a bit.
## Summary of changes
- Add a semaphore to limit how many Tenant::spawn()s may concurrently do
I/O to attach their tenant (i.e. read indices from remote storage, scan
local layer files, etc).
- Add Tenant::activate_now, a hook for kicking a tenant in its spawn()
method to skip waiting for the warmup semaphore
- For tenants that attached via warmup semaphore units, wait for logical
size calculation to complete before dropping the warmup units
- Set Tenant::activate_now in `get_active_tenant_with_timeout` (the page
service's path for getting a reference to a tenant).
- Wait for tenant activation in HTTP handlers for timeline creation and
deletion: like page service requests, these require an active tenant and
should prioritize activation if called.
## Problem
Various places in remote storage were not subject to a timeout (thereby
stuck TCP connections could hold things up), and did not respect a
cancellation token (so things like timeline deletion or tenant detach
would have to wait arbitrarily long).
## Summary of changes
- Add download_cancellable and upload_cancellable helpers, and use them
in all the places we wait for remote storage operations (with the
exception of initdb downloads, where it would not have been safe).
- Add a cancellation token arg to `download_retry`.
- Use cancellation token args in various places that were missing one
per #5066Closes: #5066
Why is this only "basic" handling?
- Doesn't express difference between shutdown and errors in return
types, to avoid refactoring all the places that use an anyhow::Error
(these should all eventually return a more structured error type)
- Implements timeouts on top of remote storage, rather than within it:
this means that operations hitting their timeout will lose their
semaphore permit and thereby go to the back of the queue for their
retry.
- Doing a nicer job is tracked in
https://github.com/neondatabase/neon/issues/6096
Part of getpage@lsn benchmark epic:
https://github.com/neondatabase/neon/issues/5771
This PR moves the control plane's spread-all-over-the-place client for
the pageserver management API into a separate module within the
pageserver crate.
I need that client to be async in my benchmarking work, so, this PR
switches to the async version of `reqwest`.
That is also the right direction generally IMO.
The switch to async in turn mandated converting most of the
`control_plane/` code to async.
Note that some of the client methods should be taking `TenantShardId`
instead of `TenantId`, but, none of the callers seem to be
sharding-aware.
Leaving that for another time:
https://github.com/neondatabase/neon/issues/6154
This doesn't make the scrubber smart enough to understand that many
shards are part of the same tenants, but it makes it understand paths
well enough to scrub the individual shards without thinking they're
malformed.
This is a prerequisite to being able to run tests with sharding enabled.
Related: #5929
## Summary of changes
saw some low-hanging codecov improvements. even if code coverage is
somewhat of a pointless game, might as well add tests where we can and
delete code if it's unused
* initdb uploads had no cancellation token, which means that when we
were stuck in upload retries, we wouldn't be able to delete the
timeline. in general, the combination of retrying forever and not having
cancellation tokens is quite dangerous.
* initdb uploads wouldn't rewind the file. this wasn't discovered in the
purposefully unreliable test-s3 in pytest because those fail on the
first byte always, not somewhere during the connection. we'd be getting
errors from the AWS sdk that the file was at an unexpected end.
slack thread: https://neondb.slack.com/archives/C033RQ5SPDH/p1702632247784079
Compaction was holding back timeline deletion because the compaction
lock had been acquired, but the semaphore was waited on. Timeline
deletion was waiting on the same lock for 1500s.
This replaces the
`pageserver::tenant::tasks::concurrent_background_tasks_rate_limit`
(which looks correct) with a simpler `..._permit` which is just an
infallible acquire, which is easier to spot "aah this needs to be raced
with cancellation tokens".
Ref: https://neondb.slack.com/archives/C03F5SM1N02/p1702496912904719
Ref: https://neondb.slack.com/archives/C03F5SM1N02/p1702578093497779
Before any json parsing from the http api only returned errors were per
field errors. Now they are done using `serde_path_to_error`, which at
least helped greatly with the `disk_usage_eviction_run` used for
testing. I don't think this can conflict with anything added in #5310.
## Problem
Historically, the pageserver used an "uninit mark" file on disk for two
purposes:
- Track which timeline dirs are incomplete for handling on restart
- Avoid trying to create the same timeline twice at the same time.
The original purpose of handling restarts is now defunct, as we use
remote storage as the source of truth and clean up any trash timeline
dirs on startup. Using the file to mutually exclude creation operations
is error prone compared with just doing it in memory, and the existing
checks happened some way into the creation operation, and could expose
errors as 500s (anyhow::Errors) rather than something clean.
## Summary of changes
- Creations are now mutually excluded in memory (using
`Tenant::timelines_creating`), rather than relying on a file on disk for
coordination.
- Acquiring unique access to the timeline ID now happens earlier in the
request.
- Creating the same timeline which already exists is now a 201: this
simplifies retry handling for clients.
- 409 is still returned if a timeline with the same ID is still being
created: if this happens it is probably because the client timed out an
earlier request and has retried.
- Colliding timeline creation requests should no longer return 500
errors
This paves the way to entirely removing uninit markers in a subsequent
change.
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
## Problem
The websockets gauge for active db connections seems to be growing more
than the gauge for client connections over websockets, which does not
make sense.
## Summary of changes
refactor how our counter-pair gauges are represented. not sure if this
will improve the problem, but it should be harder to mess-up the
counters. The API is much nicer though now and doesn't require
scopeguard::defer hacks
Dependency (commits inline):
https://github.com/neondatabase/neon/pull/5842
## Problem
Secondary mode tenants need a manifest of what to download. Ultimately
this will be some kind of heat-scored set of layers, but as a robust
first step we will simply use the set of resident layers: secondary
tenant locations will aim to match the on-disk content of the attached
location.
## Summary of changes
- Add heatmap types representing the remote structure
- Add hooks to Tenant/Timeline for generating these heatmaps
- Create a new `HeatmapUploader` type that is external to `Tenant`, and
responsible for walking the list of attached tenants and scheduling
heatmap uploads.
Notes to reviewers:
- Putting the logic for uploads (and later, secondary mode downloads)
outside of `Tenant` is an opinionated choice, motivated by:
- Enable future smarter scheduling of operations, e.g. uploading the
stalest tenant first, rather than having all tenants compete for a fair
semaphore on a first-come-first-served basis. Similarly for downloads,
we may wish to schedule the tenants with the hottest un-downloaded
layers first.
- Enable accessing upload-related state without synchronization (it
belongs to HeatmapUploader, rather than being some Mutex<>'d part of
Tenant)
- Avoid further expanding the scope of Tenant/Timeline types, which are
already among the largest in the codebase
- You might reasonably wonder how much of the uploader code could be a
generic job manager thing. Probably some of it: but let's defer pulling
that out until we have at least two users (perhaps secondary downloads
will be the second one) to highlight which bits are really generic.
Compromises:
- Later, instead of using digests of heatmaps to decide whether anything
changed, I would prefer to avoid walking the layers in tenants that
don't have changes: tracking that will be a bit invasive, as it needs
input from both remote_timeline_client and Layer.
## Problem
Single rate bucket is limited in usefulness
## Summary of changes
Introduce a secondary bucket allowing an average of 200 requests per
second over 1 minute, and a tertiary bucket allowing an average of 100
requests per second over 10 minutes.
Configured by using a format like
```sh
proxy --endpoint-rps-limit 300@1s --endpoint-rps-limit 100@10s --endpoint-rps-limit 50@1m
```
If the bucket limits are inconsistent, an error is returned on startup
```
$ proxy --endpoint-rps-limit 300@1s --endpoint-rps-limit 10@10s
Error: invalid endpoint RPS limits. 10@10s allows fewer requests per bucket than 300@1s (100 vs 300)
```
This is needed to allow use of batch queries from browsers.
## Problem
SQL-over-HTTP batch queries fail from web browsers because the relevant
headers, `Neon-Batch-isolation-Level` and `Neon-Batch-Read-Only`, are
not included in the server's OPTIONS response. I think we simply forgot
to add them when implementing the batch query feature.
## Summary of changes
Added `Neon-Batch-Isolation-Level` and `Neon-Batch-Read-Only` to the
OPTIONS response.
Changes I wanted to make on #6106 but decided to leave out to keep that
commit clean for including in the #6090. Finally remove
`PageReconstructionError::NeedsDownload`.
## Problem
Test deletes tenant and recreates with the same ID. The recreation bumps
generation number. This could lead to stale generation warnings in the
logs.
## Summary of changes
Handle this more gracefully by re-creating in the same generation that
the tenant was previously attached in.
We could also update the tenant delete path to have the attachment
service to drop tenant state on delete, but I like having it there: it
makes debug easier, and the only time it's a problem is when a test is
re-using a tenant ID after deletion.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
## Problem
1. Using chrono for durations only is wasteful
2. The arc/mutex was not being utilised
3. Locking every shard in the dashmap every GC could cause latency
spikes
4. More buckets
## Summary of changes
1. Use `Instant` instead of `NaiveTime`.
2. Remove the `Arc<Mutex<_>>` wrapper, utilising that dashmap entry
returns mut access
3. Clear only a random shard, update gc interval accordingly
4. Multiple buckets can be checked before allowing access
When I benchmarked the check function, it took on average 811ns when
multithreaded over the course of 10 million checks.
Repeated calls to `.append` don't line up as nicely as they might get
formatted in different ways. Also, it is more characters and the lines
might be longer.
Saw this while working on #5912.
- The code for calculating the prefix in the bucket was expecting a
trailing slash (as it is in the tests), but that's an awkward
expectation to impose for use in the field: make the code more flexible
by only trimming a trailing character if it is indeed a slash.
- initdb archives were detected by the scrubber as malformed layer
files. Teach it to recognize and ignore them.
## Problem
We need a reliable way to restore a project state (in this context, I
mean data on pageservers, safekeepers, and remote storage) from a
snapshot. The existing method (that we use in `test_compatibility`)
heavily relies on config files, which makes it harder to add/change
fields in the config.
The proposed solution uses config file only to get `default_tenant_id`
and `branch_name_mappings`.
## Summary of changes
- Add `NeonEnvBuilder#from_repo_dir` method, which allows using the
`neon_env_builder` fixture with data from a snapshot.
- Use `NeonEnvBuilder#from_repo_dir` in compatibility tests
Requires for https://github.com/neondatabase/neon/issues/6033
## Problem
See
https://github.com/neondatabase/company_projects/issues/111https://neondb.slack.com/archives/C03H1K0PGKH/p1700166126954079
## Summary of changes
Do not search for AUX_FILES_KEY in parent timelines
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: Arseny Sher <sher-ars@yandex.ru>
## Problem
PG16 is writing null images during relation extension.
And page server implements optimisation which replace WAL record with
FPI with page image.
So instead of WAL record ~30 bytes we store 8kb null page image.
Ans this image is almost useless, because most likely it will be shortly
rewritten with actual page content.
## Summary of changes
Do not materialize wal records with null page FPI.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
## Problem
In snowflake logs currently there is no information about the protocol,
that the client uses.
## Summary of changes
Propagate the information about the protocol together with the app_name.
In format: `{app_name}/{sql_over_http/tcp/ws}`.
This will give to @stepashka more observability on what our clients are
using.
## Problem
TenantId is changing to TenantShardId in many APIs. The swagger had
`format: hex` attributes on some of these IDs. That isn't formally
defined anywhere, but a reasonable person might think it means "hex
digits only", which will no longer be the case once we start using
shard-aware IDs (they're like `<tenant_id>-0001`).
## Summary of changes
- Remove these `format` attributes from all `tenant_id` fields in the
swagger definition
## Problem
- `shutdown_tasks` would log when a particular task was taking a long
time to shut down, but not when it eventually completed. That left one
uncertain as to whether the slow task was the source of a hang, or just
a precursor.
## Summary of changes
- Add a log line after a slow task shutdown
- Add an equivalent in Gate's `warn_if_stuck`, in case we ever need it.
This isn't related to the original issue but was noticed when checking
through these logging paths.
Error indicating request cancellation OR timeline shutdown was deemed as
a reason to exit the background worker that calculated synthetic size.
Fix it to only be considered for avoiding logging such of such errors.
These tests have been loitering on a branch of mine for a while: they
already provide value even without all the secondary mode bits landed
yet, and the Workload helper is handy for other tests too.
- `Workload` is a re-usable test workload that replaces some of the
arbitrary "write a few rows" SQL that I've found my self repeating, and
adds a systematic way to append data and check that reads properly
reflect the changes. This append+validate stuff is important when doing
migrations, as we want to detect situations where we might be reading
from a pageserver that has not properly seen latest changes.
- test_multi_attach is a validation of how the pageserver handles
attaching the same tenant to multiple pageservers, from a safety point
of view. This is intentionally separate from the larger testing of
migration, to provide an isolated environment for multi-attachment.
- test_location_conf_churn is a pseudo-random walk through the various
states that TenantSlot can be put into, with validation that attached
tenants remain externally readable when they should, and as a side
effect validating that the compute endpoint's online configuration
changes work as expected.
- test_live_migration is the reference implementation of how to drive a
pair of pageservers through a zero-downtime migration of a tenant.
---------
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
## Problem
In https://github.com/neondatabase/neon/pull/5957, the most essential
types were updated to use TenantShardId rather than TenantId. That
unblocked other work, but didn't fully enable running multiple shards
from the same tenant on the same pageserver.
## Summary of changes
- Use TenantShardId in page cache key for materialized pages
- Update mgr.rs get_tenant() and list_tenants() functions to use a shard
id, and update all callers.
- Eliminate the exactly_one_or_none helper in mgr.rs and all code that
used it
- Convert timeline HTTP routes to use tenant_shard_id
Note on page cache:
```
struct MaterializedPageHashKey {
/// Why is this TenantShardId rather than TenantId?
///
/// Usually, the materialized value of a page@lsn is identical on any shard in the same tenant. However, this
/// this not the case for certain internally-generated pages (e.g. relation sizes). In future, we may make this
/// key smaller by omitting the shard, if we ensure that reads to such pages always skip the cache, or are
/// special-cased in some other way.
tenant_shard_id: TenantShardId,
timeline_id: TimelineId,
key: Key,
}
```
## Problem
A bunch of refactorings extracted from
https://github.com/neondatabase/neon/pull/6087 (not required for it);
the most significant one is using toml instead of formatted strings.
## Summary of changes
- Use toml instead of formatted strings for config
- Skip pageserver log check if `pageserver.log` doesn't exist
- `chmod -x test_runner/regress/test_config.py`
## Problem
## Summary of changes
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
test_creating_tenant_conf_after...
- Test detaches a tenant and then re-attaches immediatel: this causes a
race between pending remote LSN update and the generation bump in the
attachment.
test_gc_cutoff:
- Test rapidly restarts a pageserver before one generation has had the
chance to process deletions from the previous generation
## Problem
The cancellation code was confusing and error prone (as seen before in
our memory leaks).
## Summary of changes
* Use the new `TaskTracker` primitve instead of JoinSet to gracefully
wait for tasks to shutdown.
* Updated libs/utils/completion to use `TaskTracker`
* Remove `tokio::select` in favour of `futures::future::select` in a
specialised `run_until_cancelled()` helper function
These changes help with identifying thrashing.
The existing `pageserver_page_cache_find_victim_iters_total` is already
useful, but, it doesn't tell us how many individual find_victim() calls
are happening, only how many clock-LRU steps happened in the entire
system,
without info about whether we needed to actually evict other data vs
just scan for a long time, e.g., because the cache is large.
The changes in this PR allows us to
1. count each possible outcome separately, esp evictions
2. compute mean iterations/outcome
I don't think anyone except me was paying close attention to
`pageserver_page_cache_find_victim_iters_total` before, so,
I think the slight behavior change of also counting iterations
for the 'iters exceeded' case is fine.
refs https://github.com/neondatabase/cloud/issues/8351
refs https://github.com/neondatabase/neon/issues/5479
## Problem
This was wasting resources: if we run a test with mock s3 we don't then
need to run it again with local fs. When we're running in CI, we don't
need to run with the mock/local storage as well as real S3. There is
some value in having CI notice/spot issues that might otherwise only
happen when running locally, but that doesn't justify the cost of
running the tests so many more times on every PR.
## Summary of changes
- For tests that used available_remote_storages or
available_s3_storages, update them to either specify no remote storage
(therefore inherit the default, which is currently local fs), or to
specify s3_storage() for the tests that actually want an S3 API.
## Problem
Currently, if one creates many shards they will all ingest all the data:
not much use! We want them to ingest a proportional share of the data
each.
Closes: #6025
## Summary of changes
- WalIngest object gets a copy of the ShardIdentity for the Tenant it
was created by.
- While iterating the `blocks` part of a decoded record, blocks that do
not match the current shard are ignored, apart from on shard zero where
they are used to update relation sizes in `observe_decoded_block` (but
not stored).
- Before committing a `DataDirModificiation` from a WAL record, we check
if it's empty, and drop the record if so. This check is necessary
(rather than just looking at the `blocks` part) because certain record
types may modify blocks in non-obvious ways (e.g.
`ingest_heapam_record`).
- Add WAL ingest metrics to record the total received, total committed,
and total filtered out
- Behaviour for unsharded tenants is unchanged: they will continue to
ingest all blocks, and will take the fast path through `is_key_local`
that doesn't bother calculating any hashes.
After this change, shards store a subset of the tenant's total data, and
accurate relation sizes are only maintained on shard zero.
---------
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
Currently using 8kB buffers, raise that to 32kB to hopefully 1/4 of
`spawn_blocking` usage. Also a drive-by fixing of last `tokio::io::copy`
to `tokio::io::copy_buf`.
There is double buffering in remote_storage and in pageserver for 8KiB
in using `tokio::io::copy` to read `BufReader<ReaderStream<_>>`.
Switches downloads and uploads to use `Stream<Item =
std::io::Result<Bytes>>`. Caller and only caller now handles setting up
buffering. For reading, `Stream<Item = ...>` is also a `AsyncBufRead`,
so when writing to a file, we now have `tokio::io::copy_buf` reading
full buffers and writing them to `tokio::io::BufWriter` which handles
the buffering before dispatching over to `tokio::fs::File`.
Additionally implements streaming uploads for azure. With azure
downloads are a bit nicer than before, but not much; instead of one huge
vec they just hold on to N allocations we got over the wire.
This PR will also make it trivial to switch reading and writing to
io-uring based methods.
Cc: #5563.
## Problem
Since #5449 we enable generations in tests by default. Running
benchmarks was missed while merging that PR, and there was one that
needed updating.
## Summary of changes
Make test_bulk_insert use the proper generation-aware helper for tenant
creation.
## Problem
Some existing tests are written in a way that's incompatible with tenant
generations.
## Summary of changes
Update all the tests that need updating: this is things like calling
through the NeonPageserver.tenant_attach helper to get a generation
number, instead of calling directly into the pageserver API. There are
various more subtle cases.
## Problem
For channel binding failed messages we were still saying "channel
binding not supported" in the errors.
## Summary of changes
Fix error messages
Per [feedback], split the Layer metrics, also finally account for lost
and [re-submitted feedback] on `layer_gc` by renaming it to
`layer_delete`, `Layer::garbage_collect_on_drop` renamed to
`Layer::delete_on_drop`. References to "gc" dropped from metric names
and elsewhere.
Also fixes how the cancellations were tracked: there was one rare
counter. Now there is a top level metric for cancelled inits, and the
rare "download failed but failed to communicate" counter is kept.
Fixes: #6027
[feedback]: https://github.com/neondatabase/neon/pull/5809#pullrequestreview-1720043251
[re-submitted feedback]: https://github.com/neondatabase/neon/pull/5108#discussion_r1401867311
The gc_timeline() function is async, but it calls the synchronous wait()
function. In the worst case, that could lead to a deadlock by using up
all tokio executor threads.
In the passing, fix a few typos in comments.
Fixes issue #6045.
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
to_string forces allocating a less than pointer sized string (costing on
stack 4 usize), using a Display formattable slug saves that. the
difference seems small, but at the same time, we log these a lot.
```
could not start the compute node: compute is in state "failed": db error: ERROR: could not access file "$libdir/timescaledb-2.10.1": No such file or directory Caused by: ERROR: could not access file "$libdir/timescaledb-2.10.1": No such file or directory
```
## Problem
See https://neondb.slack.com/archives/C04DGM6SMTM/p1700560921471619
## Summary of changes
Update relation size cache for FSM fork in WAL records filter
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
- The `Attaching tenant` log message omitted some useful information
like the generation and mode
- info-level messages about writing configuration files were
unnecessarily verbose
- During process shutdown, we don't emit logs about the various phases:
this is very cheap to log since we do it once per process lifetime, and
is helpful when figuring out where something got stuck during a hang.
## Problem
Traditionally we would detach/attach directly with curl if we wanted to
"reboot" a single tenant. That's kind of inconvenient these days,
because one needs to know a generation number to issue an attach
request.
Closes: https://github.com/neondatabase/neon/issues/6011
## Summary of changes
- Introduce a new `/reset` API, which remembers the LocationConf from
the current attachment so that callers do not have to work out the
correct configuration/generation to use.
- As an additional support tool, allow an optional `drop_cache` query
parameter, for situations where we are concerned that some on-disk state
might be bad and want to clear that as well as the in-memory state.
One might wonder why I didn't call this "reattach" -- it's because
there's already a PS->CP API of that name and it could get confusing.
In case of configuring the empty compute, API handler is waiting on
condvar for compute state change. Yet, previously if Postgres failed to
start we were just setting compute status to `Failed` without notifying.
It causes a timeout on control plane side, although we can return a
proper error from compute earlier.
With this commit API handler should be properly notified.
## Problem
When a pageserver receives a page service request identified by
TenantId, it must decide which `Tenant` object to route it to.
As in earlier PRs, this stuff is all a no-op for tenants with a single
shard: calls to `is_key_local` always return true without doing any
hashing on a single-shard ShardIdentity.
Closes: https://github.com/neondatabase/neon/issues/6026
## Summary of changes
- Carry immutable `ShardIdentity` objects in Tenant and Timeline. These
provide the information that Tenants/Timelines need to figure out which
shard is responsible for which Key.
- Augment `get_active_tenant_with_timeout` to take a `ShardSelector`
specifying how the shard should be resolved for this tenant. This mode
depends on the kind of request (e.g. basebackups always go to shard
zero).
- In `handle_get_page_at_lsn_request`, handle the case where the
Timeline we looked up at connection time is not the correct shard for
the page being requested. This can happen whenever one node holds
multiple shards for the same tenant. This is currently written as a
"slow path" with the optimistic expectation that usually we'll run with
one shard per pageserver, and the Timeline resolved at connection time
will be the one serving page requests. There is scope for optimization
here later, to avoid doing the full shard lookup for each page.
- Omit consumption metrics from nonzero shards: only the 0th shard is
responsible for tracing accurate relation sizes.
Note to reviewers:
- Testing of these changes is happening separately on the
`jcsp/sharding-pt1` branch, where we have hacked neon_local etc needed
to run a test_pg_regress.
- The main caveat to this implementation is that page service
connections still look up one Timeline when the connection is opened,
before they know which pages are going to be read. If there is one shard
per pageserver then this will always also be the Timeline that serves
page requests. However, if multiple shards are on one pageserver then
get page requests will incur the cost of looking up the correct Timeline
on each getpage request. We may look to improve this in future with a
"sticky" timeline per connection handler so that subsequent requests for
the same Timeline don't have to look up again, and/or by having postgres
pass a shard hint when connecting. This is tracked in the "Loose ends"
section of https://github.com/neondatabase/neon/issues/5507
## Problem
Panic when less than 8 bytes is presented in a startup packet.
## Summary of changes
We need there to be a 4 byte message code, so the expected min length is
8.
## Problem
It looks like because of reallocation of the buckets in previous PR, the
metric is broken in graphana.
## Summary of changes
Renamed the metric.
## Problem
In the past we've rolled out all new `compute_ctl` functionality right
to all users, which could be risky. I want to have a more fine-grained
control over what we enable, in which env and to which users.
## Summary of changes
Add an option to pass a list of feature flags to `compute_ctl`. If not
passed, it defaults to an empty list. Any unknown flags are ignored.
This allows us to release new experimental features safer, as we can
then flip the flag for one specific user, only Neon employees, free /
pro / etc. users and so on. Or control it per environment.
In the current implementation feature flags are passed via compute spec,
so they do not allow controlling behavior of `empty` computes. For them,
we can either stick with the previous approach, i.e. add separate cli
args or introduce a more generic `--features` cli argument.
Problem
-------
Before this PR, there was no concurrency limit on initial logical size
computations.
While logical size computations are lazy in theory, in practice
(production), they happen in a short timeframe after restart.
This means that on a PS with 20k tenants, we'd have up to 20k concurrent
initial logical size calculation requests.
This is self-inflicted needless overload.
This hasn't been a problem so far because the `.await` points on the
logical size calculation path never return `Pending`, hence we have a
natural concurrency limit of the number of executor threads.
But, as soon as we return `Pending` somewhere in the logical size
calculation path, other concurrent tasks get scheduled by tokio.
If these other tasks are also logical size calculations, they eventually
pound on the same bottleneck.
For example, in #5479, we want to switch the VirtualFile descriptor
cache to a `tokio::sync::RwLock`, which makes us return `Pending`, and
without measures like this patch, after PS restart, VirtualFile
descriptor cache thrashes heavily for 2 hours until all the logical size
calculations have been computed and the degree of concurrency /
concurrent VirtualFile operations is down to regular levels.
See the *Experiment* section below for details.
<!-- Experiments (see below) show that plain #5479 causes heavy
thrashing of the VirtualFile descriptor cache.
The high degree of concurrency is too much for
In the case of #5479 the VirtualFile descriptor cache size starts
thrashing heavily.
-->
Background
----------
Before this PR, initial logical size calculation was spawned lazily on
first call to `Timeline::get_current_logical_size()`.
In practice (prod), the lazy calculation is triggered by
`WalReceiverConnectionHandler` if the timeline is active according to
storage broker, or by the first iteration of consumption metrics worker
after restart (`MetricsCollection`).
The spawns by walreceiver are high-priority because logical size is
needed by Safekeepers (via walreceiver `PageserverFeedback`) to enforce
the project logical size limit.
The spawns by metrics collection are not on the user-critical path and
hence low-priority. [^consumption_metrics_slo]
[^consumption_metrics_slo]: We can't delay metrics collection
indefintely because there are TBD internal SLOs tied to metrics
collection happening in a timeline manner
(https://github.com/neondatabase/cloud/issues/7408). But let's ignore
that in this issue.
The ratio of walreceiver-initiated spawns vs
consumption-metrics-initiated spawns can be reconstructed from logs
(`spawning logical size computation from context of task kind {:?}"`).
PR #5995 and #6018 adds metrics for this.
First investigation of the ratio lead to the discovery that walreceiver
spawns 75% of init logical size computations.
That's because of two bugs:
- In Safekeepers: https://github.com/neondatabase/neon/issues/5993
- In interaction between Pageservers and Safekeepers:
https://github.com/neondatabase/neon/issues/5962
The safekeeper bug is likely primarily responsible but we don't have the
data yet. The metrics will hopefully provide some insights.
When assessing production-readiness of this PR, please assume that
neither of these bugs are fixed yet.
Changes In This PR
------------------
With this PR, initial logical size calculation is reworked as follows:
First, all initial logical size calculation task_mgr tasks are started
early, as part of timeline activation, and run a retry loop with long
back-off until success. This removes the lazy computation; it was
needless complexity because in practice, we compute all logical sizes
anyways, because consumption metrics collects it.
Second, within the initial logical size calculation task, each attempt
queues behind the background loop concurrency limiter semaphore. This
fixes the performance issue that we pointed out in the "Problem" section
earlier.
Third, there is a twist to queuing behind the background loop
concurrency limiter semaphore. Logical size is needed by Safekeepers
(via walreceiver `PageserverFeedback`) to enforce the project logical
size limit. However, we currently do open walreceiver connections even
before we have an exact logical size. That's bad, and I'll build on top
of this PR to fix that
(https://github.com/neondatabase/neon/issues/5963). But, for the
purposes of this PR, we don't want to introduce a regression, i.e., we
don't want to provide an exact value later than before this PR. The
solution is to introduce a priority-boosting mechanism
(`GetLogicalSizePriority`), allowing callers of
`Timeline::get_current_logical_size` to specify how urgently they need
an exact value. The effect of specifying high urgency is that the
initial logical size calculation task for the timeline will skip the
concurrency limiting semaphore. This should yield effectively the same
behavior as we had before this PR with lazy spawning.
Last, the priority-boosting mechanism obsoletes the `init_order`'s grace
period for initial logical size calculations. It's a separate commit to
reduce the churn during review. We can drop that commit if people think
it's too much churn, and commit it later once we know this PR here
worked as intended.
Experiment With #5479
---------------------
I validated this PR combined with #5479 to assess whether we're making
forward progress towards asyncification.
The setup is an `i3en.3xlarge` instance with 20k tenants, each with one
timeline that has 9 layers.
All tenants are inactive, i.e., not known to SKs nor storage broker.
This means all initial logical size calculations are spawned by
consumption metrics `MetricsCollection` task kind.
The consumption metrics worker starts requesting logical sizes at low
priority immediately after restart. This is achieved by deleting the
consumption metrics cache file on disk before starting
PS.[^consumption_metrics_cache_file]
[^consumption_metrics_cache_file] Consumption metrics worker persists
its interval across restarts to achieve persistent reporting intervals
across PS restarts; delete the state file on disk to get predictable
(and I believe worst-case in terms of concurrency during PS restart)
behavior.
Before this patch, all of these timelines would all do their initial
logical size calculation in parallel, leading to extreme thrashing in
page cache and virtual file cache.
With this patch, the virtual file cache thrashing is reduced
significantly (from 80k `open`-system-calls/second to ~500
`open`-system-calls/second during loading).
### Critique
The obvious critique with above experiment is that there's no skipping
of the semaphore, i.e., the priority-boosting aspect of this PR is not
exercised.
If even just 1% of our 20k tenants in the setup were active in
SK/storage_broker, then 200 logical size calculations would skip the
limiting semaphore immediately after restart and run concurrently.
Further critique: given the two bugs wrt timeline inactive vs active
state that were mentioned in the Background section, we could have 75%
of our 20k tenants being (falsely) active on restart.
So... (next section)
This Doesn't Make Us Ready For Async VirtualFile
------------------------------------------------
This PR is a step towards asynchronous `VirtualFile`, aka, #5479 or even
#4744.
But it doesn't yet enable us to ship #5479.
The reason is that this PR doesn't limit the amount of high-priority
logical size computations.
If there are many high-priority logical size calculations requested,
we'll fall over like we did if #5479 is applied without this PR.
And currently, at very least due to the bugs mentioned in the Background
section, we run thousands of high-priority logical size calculations on
PS startup in prod.
So, at a minimum, we need to fix these bugs.
Then we can ship #5479 and #4744, and things will likely be fine under
normal operation.
But in high-traffic situations, overload problems will still be more
likely to happen, e.g., VirtualFile cache descriptor thrashing.
The solution candidates for that are orthogonal to this PR though:
* global concurrency limiting
* per-tenant rate limiting => #5899
* load shedding
* scaling bottleneck resources (fd cache size (neondatabase/cloud#8351),
page cache size(neondatabase/cloud#8351), spread load across more PSes,
etc)
Conclusion
----------
Even with the remarks from in the previous section, we should merge this
PR because:
1. it's an improvement over the status quo (esp. if the aforementioned
bugs wrt timeline active / inactive are fixed)
2. it prepares the way for
https://github.com/neondatabase/neon/pull/6010
3. it gets us close to shipping #5479 and #4744
# Problem
I need walredo to be cancellation-safe for
https://github.com/neondatabase/neon/pull/6000#discussion_r1412049728
# Solution
We are only `async fn` because of
`wait_for(stderr_logger_task_done).await`, added in #5560 .
The `stderr_logger_cancel` and `stderr_logger_task_done` were there out
of precaution that the stderr logger task might for some reason not stop
when the walredo process terminates.
That hasn't been a problem in practice.
So, simplify things:
- remove `stderr_logger_cancel` and the
`wait_for(...stderr_logger_task_done...)`
- use `tokio::process::ChildStderr` in the stderr logger task
- add metrics to track number of running stderr logger tasks so in case
I'm wrong here, we can use these metrics to identify the issue (not
planning to put them into a dashboard or anything)
## Problem
Limit the number of open connections between the control plane and
proxy.
## Summary of changes
Enable dynamic rate limiter in prod.
Unfortunately the latency metrics are a bit broken, but from logs I see
that on staging for the past 7 days only 2 times latency for acquiring
was greater than 1ms (for most of the cases it's insignificant).
If there are too many L0 layers before compaction, the compaction
process becomes slow because of slow `Timeline::get`. As a result of the
slowdown, the pageserver will generate even more L0 layers for the next
iteration, further exacerbating the slow performance.
Change to perform L0 -> L1 compaction before creating new images. The
simple change speeds up compaction time and `Timeline::get` to 5x.
`Timeline::get` is faster on top of L1 layers.
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
## Problem
`test_pg_clients` reads the actual result from a *.stdout file,
https://github.com/neondatabase/neon/pull/5977 has added a header to
such files, so `test_pg_clients` started to fail.
## Summary of changes
- Use `capture_stdout` and compare the expected result with the output
instead of *.stdout file content
## Problem
On pageservers upgraded to enable generations, these INFO level logs
were rather frequent. If a tenant timeline hasn't written new layers
since the upgrade, it will emit the "No index_part.json*" log every time
it starts.
## Summary of changes
- Downgrade two log lines from info to debug
- Add a tiny unit test that I wrote for sanity-checking that there
wasn't something wrong with our Generation-comparing logic when loading
index parts.
Temporarily reset neon.max_cluster_size to avoid
the possibility of hitting the limit, while we are applying config:
creating new extensions, roles, etc...
Only introduced a few hours ago (#5995), I took a look at the numbers
from staging and realized that `get_current_logical_size()` is on the
walingest hot path: we call it for every `ReplicationMessage::XLogData`
that we receive.
Since the metric is global, it would be quite a busy cache line.
This PR replaces it with a new metric purpose-built for what's most
interesting right now.
It was using `PRECONDITION_FAILED` for errors during `ComputeSpec` to
`ParsedSpec` conversion, but this disobeys the OpenAPI spec [1] and
correct code should be `BAD_REQUEST` for any spec processing errors.
While on it, I also noticed that `compute_ctl` OpenAPI spec has an
invalid format and fixed it.
[1] fd81945a60/compute_tools/src/http/openapi_spec.yaml (L119-L120)
## Problem
While investigating https://github.com/neondatabase/neon/issues/5854, we
hypothesised that logs/repo-dir from the initial failure might leak into
reruns. Use different directories for each run to avoid such a
possibility.
## Summary of changes
- make each test rerun use different directories
- update `pytest-rerunfailure` plugin from 11.1.2 to 13.0
I would love to not expose the in-accurate value int he mgmt API at all,
and in fact control plane doesn't use it [^1].
But our tests do, and I have no desire to change them at this time.
[^1]: https://github.com/neondatabase/cloud/pull/8317
... by panicking on overflow;
It was made fallible initially due to in-confidence in logical size
calculation. However, the error has never happened since I am at Neon.
Let's stop worrying about this by converting the overflow check into a panic.
If `index_part.json` is (verifiably) not present on remote storage, we
should regard the timeline as inexistent. This lets `clean_up_timelines`
purge the partial local disk state, which is important in the case of
incomplete creations leaving behind state that hinders retries. For
incomplete deletions, we also want the timeline's local disk content be
gone completely.
The PR removes the allowed warnings added by #5390 and #5912, as we now
are only supposed to issue info level messages. It also adds a
reproducer for #6007, by parametrizing the
`test_timeline_init_break_before_checkpoint_recreate` test added by
#5390. If one reverts the .rs changes, the "cannot create its uninit
mark file" log line occurs once one comments out the failing checks for
the local disk state being actually empty.
Closes#6007
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
This PR adds an `existing_initdb_timeline_id` option to timeline
creation APIs, taking an optional timeline ID.
Follow-up of #5390.
If the `existing_initdb_timeline_id` option is specified via the HTTP
API, the pageserver downloads the existing initdb archive from the given
timeline ID and extracts it, instead of running initdb itself.
---------
Co-authored-by: Christian Schwarz <christian@neon.tech>
## Problem
no problem
## Summary of changes
replaces boxstr with arcstr as it's cheaper to clone. mild perf
improvement.
probably should look into other smallstring optimsations tbh, they will
likely be even better. The longest endpoint name I was able to construct
is something like `ep-weathered-wildflower-12345678` which is 32 bytes.
Most string optimisations top out at 23 bytes
aiming for faster to understand a bunch of `.stdout` and `.stderr`
files, see example echo_1.stdout differences:
```
+# echo foobar abbacd
+
foobar abbacd
```
it can be disabled and is disabled in this PR for some tests; use
`pg_bin.run_capture(..., with_command_header=False)` for that.
as a bonus this cleans up the echoed newlines from s3_scrubber
output which are also saved to file but echoed to test log.
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
we have test cases which launch processes from threads, and they capture
output assuming this counter is thread-safe. at least according to my
understanding this operation in python requires a lock to be
thread-safe.
- During migration of tenants, it is useful for callers to
`/location_conf` to flush a tenant's layers while transitioning to
AttachedStale: this optimization reduces the redundant WAL replay work
that the tenant's new attached pageserver will have to do. Test coverage
for this will come as part of the larger tests for live migration in
#5745#5842
- Flushing is controlled with `flush_ms` query parameter: it is the
caller's job to decide how long they want to wait for a flush to
complete. If flush is not complete within the time limit, the pageserver
proceeds to succeed anyway: flushing is only an optimization.
- Add swagger definitions for all this: the location_config API is the
primary interface for driving tenant migration as described in
docs/rfcs/028-pageserver-migration.md, and will eventually replace the
various /attach /detach /load /ignore APIs.
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
Test failure seen here:
https://neon-github-public-dev.s3.amazonaws.com/reports/pr-5860/7032903218/index.html#suites/837740b64a53e769572c4ed7b7a7eeeb/c0f1c79a70a3b9ab
```
E AssertionError: assert not [(302, '2023-11-29T13:23:51.046801Z ERROR request{method=PUT path=/v1/tenant/f6b845de60cb0e92f4426e0d6af1d2ea/timeline/69a8c98004abe71a281cff8642a45274/checkpoint request_id=eca33d8a-7af2-46e7-92ab-c28629feb42c}: Error processing HTTP request: InternalServerError(queue is in state Stopped\n')]
```
This appears to be a legitimate log: the test is issuing checkpoint
requests in the background, and deleting (therefore shutting down) a
timeline.
## Problem
fixes#5654 - WakeComputeErrors occuring during a connect_to_compute got
propagated as IO errors, which get forwarded to the user as "Couldn't
connect to compute node" with no helpful message.
## Summary of changes
Handle WakeComputeError during ConnectionError properly
Enable the pageserver to recover from data corruption events by
implementing a feature to re-apply historic WAL records in parallel to the already
occurring WAL replay.
The feature is outside of the user-visible backup and history story, and
only
serves as a second-level backup for the case that there is a bug in the
pageservers that corrupted the served pages.
The RFC proposes the addition of two new features:
* recover a broken branch from WAL (downtime is allowed)
* a test recovery system to recover random branches to make sure
recovery works
## Problem
Per-project IP allowlist:
https://github.com/neondatabase/cloud/issues/8116
## Summary of changes
Implemented IP filtering on the proxy side.
To retrieve ip allowlist for all scenarios, added `get_auth_info` call
to the control plane for:
* sql-over-http
* password_hack
* cleartext_hack
Added cache with ttl for sql-over-http path
This might slow down a bit, consider using redis in the future.
---------
Co-authored-by: Conrad Ludgate <conrad@neon.tech>
Remove handcrafted TenantConf deserialization code. Use
`serde_path_to_error` to include the field which failed parsing. Leaves
the duplicated TenantConf in pageserver and models, does not touch
PageserverConf handcrafted deserialization.
Error change:
- before change: "configure option `checkpoint_distance` cannot be
negative"
- after change: "`checkpoint_distance`: invalid value: integer `-1`,
expected u64"
Fixes: #5300
Cc: #3682
---------
Signed-off-by: Rahul Modpur <rmodpur2@gmail.com>
Co-authored-by: Shany Pozin <shany@neon.tech>
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
## Problem
#5123
## Summary of changes
Add `--sql-over-http-pool-opt-in true` default cli arg. Allows us to set
`--sql-over-http-pool-opt-in false` region-by-region
`neon_local endpoint` subcommand currently allows creating two primary
endpoints for the same branch which leads to shutdown of both endpoints
`neon_local endpoint start` new behavior:
1. Fail if endpoint doesn't exist
2. Fail if two primary conflict detected
Fixes#4959Closes#5426
Signed-off-by: Rahul Modpur <rmodpur2@gmail.com>
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
(includes two preparatory commits from
https://github.com/neondatabase/neon/pull/5960)
## Problem
To accommodate multiple shards in the same tenant on the same
pageserver, we must include the full TenantShardId in local paths. That
means that all code touching local storage needs to see the
TenantShardId.
## Summary of changes
- Replace `tenant_id: TenantId` with `tenant_shard_id: TenantShardId` on
Tenant, Timeline and RemoteTimelineClient.
- Use TenantShardId in helpers for building local paths.
- Update all the relevant call sites.
This doesn't update absolutely everything: things like PageCache,
TaskMgr, WalRedo are still shard-naive. The purpose of this PR is to
update the core types so that others code can be added/updated
incrementally without churning the most central shared types.
Precursor for https://github.com/neondatabase/neon/pull/5957
## Problem
When DeletionList was written, TenantId/TimelineId didn't have
human-friendly modes in their serde. #5335 added those, such that the
helpers used in serialization of HashMaps are no longer necessary.
## Summary of changes
- Add a unit test to ensure that this change isn't changing anything
about the serialized form
- Remove the serialization helpers for maps of Id
## Problem
PG16 (https://github.com/neondatabase/postgres/pull/327) adds new
function to SMGR: zeroextend
It's implementation in Neon actually wal-log zero pages of extended
relation.
This zero page is wal-logged using XLOG_FPI.
As far as page is zero, the hole optimization (excluding from the image
everything between pg_upper and pd_lower) doesn't work.
## Summary of changes
In case of zero page (`PageIsNull()` returns true) assume
`hole_size=BLCKSZ`
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
## Problem
This is a narrow race that can leave a stuck Stopping tenant behind,
while emitting a log error "Missing InProgress marker during tenant
upsert, this is a bug"
- Deletion request 1 puts tenant into Stopping state, and fires off
background part of DeleteTenantFlow
- Deletion request 2 acquires a SlotGuard for the same tenant ID, leaves
a TenantSlot::InProgress in place while it checks if the tenant's state
is accept able.
- DeleteTenantFlow finishes, calls TenantsMap::remove, which removes the
InProgress marker.
- Deletion request 2 calls SlotGuard::revert, which upserts the old
value (the Tenant in Stopping state), and emits the telltale log
message.
Closes: #5936
## Summary of changes
- Add a regression test which uses pausable failpoints to reproduce this
scenario.
- TenantsMap::remove is only called by DeleteTenantFlow. Its behavior is
tweaked to express the different possible states, especially
`InProgress` which carriers a barrier.
- In DeleteTenantFlow, if we see such a barrier result from remove(),
wait for the barrier and then try removing again.
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
pgbouncer 1.21.0 doesn't play nicely with DEALLOCATD/DISCARD ALL if
prepared statement support is enabled (max_prepared_statements > 0).
There's a patch[0] that improves this (it will be included in the next
release of pgbouncer).
This PR applies this patch on top of 1.21.0 release tarball.
For some reason, the tarball doesn't include `test/test_prepared.py`
(which is modified by the patch as well), so the patch can't be applied
clearly. I use `filterdiff` (from `patchutils` package) to apply
the required changes.
[0] a7b3c0a5f4
Quest: https://github.com/neondatabase/neon/issues/4745. Follow-up to
#4938.
- add in locks for compaction and gc, so we don't have multiple
executions at the same time in tests
- remove layer_removal_cs
- remove waiting for uploads in eviction/gc/compaction
- #4938 will keep the file resident until upload completes
Co-authored-by: Christian Schwarz <christian@neon.tech>
## Problem
The TenantShardId in API URLs is sufficient to uniquely identify a
tenant shard, but not for it to function: it also needs to know its full
sharding configuration (stripe size, layout version) in order to map
keys to shards.
## Summary of changes
- Introduce ShardIdentity: this is the superset of ShardIndex (#5924 )
that is required for translating keys to shard numbers.
- Include ShardIdentity as an optional attribute of LocationConf
- Extend the public `LocationConfig` API structure with a flat
representation of shard attributes.
The net result is that at the point we construct a `Tenant`, we have a
`ShardIdentity` (inside LocationConf). This enables the next steps to
actually use the ShardIdentity to split WAL and validate that page
service requires are reaching the correct shard.
## Problem
For sharded tenants, the layer keys must include the shard number and
shard count, to disambiguate keys written by different shards in the
same tenant (shard number), and disambiguate layers written before and
after splits (shard count).
Closes: https://github.com/neondatabase/neon/issues/5924
## Summary of changes
There are no functional changes in this PR: everything behaves the same
for the default ShardIndex::unsharded() value. Actual construct of
sharded tenants will come next.
- Add a ShardIndex type: this is just a wrapper for a ShardCount and
ShardNumber. This is a subset of ShardIdentity: whereas ShardIdentity
contains enough information to filter page keys, ShardIndex contains
just enough information to construct a remote key. ShardIndex has a
compact encoding, the same as the shard part of TenantShardId.
- Store the ShardIndex as part of IndexLayerMetadata, if it is set to a
different value than ShardIndex::unsharded.
- Update RemoteTimelineClient and DeletionQueue to construct paths using
the layer metadata. Deletion code paths that previously just passed a
`Generation` now pass a full `LayerFileMetadata` to capture the shard as
well.
Notes to reviewers:
- In deletion code paths, I could have used a (Generation, ShardIndex)
instead of the full LayerFileMetadata. I opted for the full object
partly for brevity, and partly because in future when we add checksums
the deletion code really will care about the full metadata in order to
validate that it is deleting what was intended.
- While ShardIdentity and TenantShardId could both use a ShardIndex, I
find that they read more cleanly as "flat" structs that spell out the
shard count and number field separately. Serialization code would need
writing out by hand anyway, because TenantShardId's serialized form is
not a serde struct-style serialization.
- ShardIndex doesn't _have_ to exist (we could use ShardIdentity
everywhere), but it is a worthwhile optimization, as we will have many
copies of this as part of layer metadata. In future the size difference
betweedn ShardIndex and ShardIdentity may become larger if we implement
more sophisticated key distribution mechanisms (i.e. new values of
ShardIdentity::layout).
---------
Co-authored-by: Christian Schwarz <christian@neon.tech>
## Problem
We can segfault if we update connstr inside of a process that has
detached from shmem (e.g. inside stats collector)
## Summary of changes
Add a check to make sure we're not detached
## Problem
channel binding protects scram from sophisticated MITM attacks where the
attacker is able to produce 'valid' TLS certificates.
## Summary of changes
get the tls-server-end-point channel binding, and verify it is correct
for the SCRAM-SHA-256-PLUS authentication flow
instead of direct S3 request.
Pros:
- simplify code a lot (no need to provide AWS credentials and paths);
- reduce latency of downloading extension data as proxy resides near
computes; -reduce AWS costs as proxy has cache and 1000 computes asking
the same extension will not generate 1000 downloads from S3.
- we can use only one S3 bucket to store extensions (and rid of regional
buckets which were introduced to reduce latency);
Changes:
- deprecate remote-ext-config compute_ctl parameter, use
http://pg-ext-s3-gateway if any old format remote-ext-cofig is provided;
- refactor tests to use mock http server;
## Problem
Historically, we treated the presence of a timeline on local disk as
evidence that it logically exists. Since #5580 that is no longer the
case, so we can always rely on remote storage. If we restart and the
timeline is gone in remote storage, we will also purge it from local
disk: no need for a marker.
Reference on why this PR is for timeline markers and not tenant markers:
https://github.com/neondatabase/neon/issues/5080#issuecomment-1783187807
## Summary of changes
Remove code paths that read + write deletion marker for timelines.
Leave code path that deletes these markers, just in case we deploy while
there are some in existence. This can be cleaned up later.
(https://github.com/neondatabase/neon/issues/5718)
Fix the fallible vs. infallible check order with
`UninitTimeline::finish_creation` so that the incomplete timeline can be
removed. Currently the order of drop guard unwrapping causes uninit
files to be left on pageserver, blocking the tenant deletion.
Cc: #5914
Cc: #investigation-2023-11-23-stuck-tenant-deletion
this will make it easier to test if an added allowed_error does in fact
match for example against a log file from an allure report.
```
$ python3 test_runner/fixtures/pageserver/allowed_errors.py --help
usage: allowed_errors.py [-h] [-i INPUT]
check input against pageserver global allowed_errors
optional arguments:
-h, --help show this help message and exit
-i INPUT, --input INPUT
Pageserver logs file. Reads from stdin if no file is provided.
```
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
## Problem
There is not check that LFC is initialised (`lfc_max_size != 0`) in
`local_cache_pages` function
## Summary of changes
Add proper check.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
## Problem
The connections.join_next helped but it wasn't enough... The way I
implemented the improvement before was still faulty but it mostly worked
so it looked like it was working correctly.
From [`tokio::select`
docs](https://docs.rs/tokio/latest/tokio/macro.select.html):
> 4. Once an <async expression> returns a value, attempt to apply the
value to the provided <pattern>, if the pattern matches, evaluate
<handler> and return. If the pattern does not match, disable the current
branch and for the remainder of the current call to select!. Continue
from step 3.
The `connections.join_next()` future would complete and `Some(Err(e))`
branch would be evaluated but not match (as the future would complete
without panicking, we would hope). Since the branch doesn't match, it's
disabled. The select continues but never attempts to call `join_next`
again. Getting unlucky, more TCP connections are created than we attempt
to join_next.
## Summary of changes
Replace the `Some(Err(e))` pattern with `Some(e)`. Because of the
auto-disabling feature, we don't need the `if !connections.is_empty()`
step as the `None` pattern will disable it for us.
- Run CREATE EXTENSION neon for template1, so that it was created in all databases.
- Run ALTER EXTENSION neon in all databases, to always have the newest version of the extension in computes.
- Add test_neon_extension test
## Problem
See #2592
## Summary of changes
Compresses the results of initdb into a .tar.zst file and uploads them
to S3, to enable usage in recovery from lsn.
Generations should not be involved I think because we do this only once
at the very beginning of a timeline.
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
## Problem
#5900
## Summary of changes
Added cancellation token as param in all relevant code paths and actually used it in the find_lsn_for_timestamp main loop
fixes https://github.com/neondatabase/neon/issues/5878
obsoletes https://github.com/neondatabase/neon/issues/5879
Before this PR, it could happen that `load_layer_map` schedules removal
of the future
image layer. Then a later compaction run could re-create the same image
layer, scheduling a PUT.
Due to lack of an upload queue barrier, the PUT and DELETE could be
re-ordered.
The result was IndexPart referencing a non-existent object.
## Summary of changes
* Add support to `pagectl` / Python tests to decode `IndexPart`
* Rust
* new `pagectl` Subcommand
* `IndexPart::{from,to}_s3_bytes()` methods to internalize knowledge
about encoding of `IndexPart`
* Python
* new `NeonCli` subclass
* Add regression test
* Rust
* Ability to force repartitioning; required to ensure image layer
creation at last_record_lsn
* Python
* The regression test.
* Fix the issue
* Insert an `UploadOp::Barrier` after scheduling the deletions.
## Problem
## Summary of changes
See #5500
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
(part of the getpage benchmarking epic #5771)
The plan is to make the benchmarking tool log on stderr and emit results
as JSON on stdout. That way, the test suite can simply take captures
stdout and json.loads() it, while interactive users of the benchmarking
tool have a reasonable experience as well.
Existing logging users continue to print to stdout, so, this change
should be a no-op functionally and performance-wise.
This way, `cargo update -p tokio-postgres` just works. The `Cargo.toml`
communicates more clearly that we're referring to the `main` branch. And
the git revision is still pinned in `Cargo.lock`.
While reviewing code noticed a scary `layer_paths.pop().unwrap()` then
realized this should be further asyncified, something I forgot to do
when I switched the `compact_level0_phase1` back to async in #4938.
This keeps the double-fsync for new deltas as #4749 is still unsolved.
## Problem
Currently, control plane doesn't know about neon_superuser, so if a user
creates a database with owner neon_superuser it causes an exception when
it tries to forward it. It is also currently possible to ALTER ROLE
neon_superuser.
## Summary of changes
Disallow creating database with owner neon_superuser. This is probably
fine, since I don't think you can create a database with owner normal
superuser. Also forbids altering neon_superuser
This change brings down incremental compilation for me
from > 1min to 10s (and this is a pretty old Ryzen 1700X).
More details: "incremental compilation" here means to change one
character
in the `failed to read value from offset` string in `image_layer.rs`.
The command for incremental compilation is `cargo build_testing`.
The system on which I got these numbers uses `mold` via
`~/.cargo/config.toml`.
As a bonus, `rust-gdb` is now at least a little fun again.
Some tests are timing out in debug builds due to these changes.
This PR makes them skip for debug builds.
We run both with debug and release build, so, the loss of coverage is
marginal.
---------
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
A very low number of layer loads have been marked wrongly as permanent,
as I did not remember that `VirtualFile::open` or reading could fail
transiently for contention. Return separate errors for transient and
persistent errors from `{Delta,Image}LayerInner::load`.
Includes drive-by comment changes.
The implementation looks quite ugly because having the same type be both
the inner (operation error) and outer (critical error), but with the
alternatives I tried I did not find a better way.
The longer a pageserver runs, the more walredo processes it accumulates
from tenants that are touched intermittently (e.g. by availability
checks). This can lead to getting OOM killed.
Changes:
- Add an Instant recording the last use of the walredo process for a
tenant
- After compaction iteration in the background task, check for idleness
and stop the walredo process if idle for more than 10x compaction
period.
Cc: #3620
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
Co-authored-by: Shany Pozin <shany@neon.tech>
First compaction/gc error backoff starts from 0 which is less than 2s
what it was before #5672. This is now fixed to be the intended 2**n.
Additionally noticed the `compaction_iteration` creating an
`anyhow::Error` via `into()` always captures a stacktrace even if we had
a stacktraceful anyhow error within the CompactionError because there is
no stable api for querying that.
Changes the error message encountered when the `neon.max_cluster_size`
limit is reached. Reasoning is that this is user-visible, and so should
*probably* use language that's closer to what users are familiar with.
## Problem
We don't know the number of users with the different kind of
authentication: ["sni", "endpoint in options" (A and B from
[here](https://neon.tech/docs/connect/connection-errors)),
"password_hack"]
## Summary of changes
Collect metrics by sni kind.
## Problem
When using TenantId as the key, we are unable to handle multiple tenant
shards attached to the same pageserver for the same tenant ID. This is
an expected scenario if we have e.g. 8 shards and 5 pageservers.
## Summary of changes
- TenantsMap is now a BTreeMap instead of a HashMap: this enables
looking up by range. In future, we will need this for page_service, as
incoming requests will just specify the Key, and we'll have to figure
out which shard to route it to.
- A new key type TenantShardId is introduced, to act as the key in
TenantsMap, and as the id type in external APIs. Its human readable
serialization is backward compatible with TenantId, and also
forward-compatible as long as sharding is not actually used (when we
construct a TenantShardId with ShardCount(0), it serializes to an
old-fashioned TenantId).
- Essential tenant APIs are updated to accept TenantShardIds:
tenant/timeline create, tenant delete, and /location_conf. These are the
APIs that will enable driving sharded tenants. Other apis like /attach
/detach /load /ignore will not work with sharding: those will soon be
deprecated and replaced with /location_conf as part of the live
migration work.
Closes: #5787
## Problem
`HEAD` commit for a PR is a phantom merge commit which skews the baseline
commit for coverage reports.
See
https://github.com/neondatabase/neon/pull/5751#issuecomment-1790717867
## Summary of changes
- Use commit hash instead of `HEAD` for finding baseline commits for
code coverage
- Use the base branch for PRs or the current branch for pushes
- trivial serialization roundtrip test for
`pageserver::repository::Value`
- add missing `start_paused = true` to 15s test making it <0s test
- completely unrelated future clippy lint avoidance (helps beta channel
users)
## Problem
Currently the only way to exercise tenant migration is via python test
code. We need a convenient way for developers to do it directly in a
neon local environment.
## Summary of changes
- Add a `--num-pageservers` argument to `cargo neon init` so that it's
easy to run with multiple pageservers
- Modify default pageserver overrides in neon_local to set up `LocalFs`
remote storage, as any migration/attach/detach stuff doesn't work in the
legacy local storage mode. This also unblocks removing the pageserver's
support for the legacy local mode.
- Add a new `cargo neon tenant migrate` command that orchestrates tenant
migration, including endpoints.
`cargo deny` was complaining the version 0.8.3 was yanked (for possible
DoS attack [wiki]), but the latest version (0.8.5) also includes aarch64
fixes which may or may not be relevant. Our usage of ahash limits to
proxy, but I don't think we are at any risk.
[wiki]: https://github.com/tkaitchuck/aHash/wiki/Yanked-versions
rationale: some crates pull these in as default; hopefully these hints
will require less cleanup-after and Cargo.lock file watching.
follow-up to #5848.
## Problem
This is a log hygiene fix, for an occasional test failure.
warn-level logging in imitate_timeline_cached_layer_accesses can't
distinguish actual errors from shutdown cases.
## Summary of changes
Replaced anyhow::Error with an explicit CollectKeySpaceError type, that
includes conversion from PageReconstructError::Cancelled.
In OOM situations, knowing exactly how many walredo processes there were
at a time would help afterwards to understand why was pageserver OOM
killed. Add `pageserver_wal_redo_process_total` metric to keep track of
total wal redo process started, shutdown and killed since pageserver
start.
Closes#5722
---------
Signed-off-by: Rahul Modpur <rmodpur2@gmail.com>
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
Co-authored-by: Christian Schwarz <me@cschwarz.com>
## Problem
I've added `BUILD_TAG` to docker images.
(https://github.com/neondatabase/neon/pull/5812), but forgot to add it
to services that we build for tests
## Summary of changes
- Set `BUILD_TAG` in `build-neon` job
## Problem
Some developers use workstations with arm CPUs, and sometimes x86-64
code is not fully compatible with it (for example,
https://github.com/neondatabase/neon/pull/5827).
Although we don't have arm CPUs in the prod (yet?), it is worth having
some basic checks for this architecture to have a better developer
experience.
Closes https://github.com/neondatabase/neon/issues/5829
## Summary of changes
- Run `check-codestyle-rust`-like & `build-neon`-like jobs on Arm runner
- Add `run-extra-build-*` label to run all available extra builds
Includes the changes of #3689 that address point 1 of #3689, plus some
further improvements. In particular, this PR does:
* set `min_lsn` to a safe value to create branches from (and verify it
in tests)
* return `min_lsn` instead of `max_lsn` for `NoData` and `Past` (verify
it in test for `Past`, `NoData` is harder and not as important)
* return `commit_lsn` instead of `max_lsn` for Future (and verify it in
the tests)
* add some comments
Split out of #5686 to get something more minimal out to users.
Minor bugfix, something noticed while manual code-review. Use the same
joinset for inprogress tenants so we can get the benefit of the
buffering logging just as we get for attached tenants, and no single
inprogress task can hold up shutdown of other tenants.
The `get_timestamp_of_lsn` pageserver endpoint has been added in #5497,
but the yml it added was wrong: the lsn is expected in hex format, not
in integer (decimal) format.
## Problem
We have observed the shutdown of a timeline taking a long time when a
deletion arrives at a busy time for the system. This suggests that we
are not respecting cancellation tokens promptly enough.
## Summary of changes
- Refactor timeline shutdown so that rather than having a shutdown()
function that takes a flag for optionally flushing, there are two
distinct functions, one for graceful flushing shutdown, and another that
does the "normal" shutdown where we're just setting a cancellation token
and then tearing down as fast as we can. This makes things a bit easier
to reason about, and enables us to remove the hand-written variant of
shutdown that was maintained in `delete.rs`
- Layer flush task checks cancellation token more carefully
- Logical size calculation's handling of cancellation tokens is
simplified: rather than passing one in, it respects the Timeline's
cancellation token.
This PR doesn't touch RemoteTimelineClient, which will be a key thing to
fix as well, so that a slow remote storage op doesn't hold up shutdown.
replace the checkbox list with a a proper task list in the epic template
NB: this PR does not change the code, it only touches the github issue
templates
## Problem
A user can perform many database connections at the same instant of time
- these will all cache miss and materialise as requests to the control
plane. #5705
## Summary of changes
I am using a `DashMap` (a sharded `RwLock<HashMap>`) of endpoints ->
semaphores to apply a limiter. If the limiter is enabled (permits > 0),
the semaphore will be retrieved per endpoint and a permit will be
awaited before continuing to call the wake_compute endpoint.
### Important details
This dashmap would grow uncontrollably without maintenance. It's not a
cache so I don't think an LRU-based reclamation makes sense. Instead,
I've made use of the sharding functionality of DashMap to lock a single
shard and clear out unused semaphores periodically.
I ran a test in release, using 128 tokio tasks among 12 threads each
pushing 1000 entries into the map per second, clearing a shard every 2
seconds (64 second epoch with 32 shards). The endpoint names were
sampled from a gamma distribution to make sure some overlap would occur,
and each permit was held for 1ms. The histogram for time to clear each
shard settled between 256-512us without any variance in my testing.
Holding a lock for under a millisecond for 1 of the shards does not
concern me as blocking
## Problem
Currently, we aren't doing any explicit slowdown in response to 429
responses. Recently, as we hit remote storage a bit harder (pageserver
does more ListObjectsv2 requests than it used to since #5580 ), we're
seeing storms of 429 responses that may be the result of not just doing
too may requests, but continuing to do those extra requests without
backing off any more than our usual backoff::exponential.
## Summary of changes
Switch from AWS's "Standard" retry policy to "Adaptive" -- docs describe
this as experimental but it has been around for a long time. The main
difference between Standard and Adaptive is that Adaptive rate-limits
the client in response to feedback from the server, which is meant to
avoid scenarios where the client would otherwise repeatedly hit
throttling responses.
## Problem
We have a funny 3-day timeout for connections between the compute and
pageserver. We want to get rid of it, so to do that we need to make sure
the compute is resilient to connection failures.
Closes: https://github.com/neondatabase/neon/issues/5518
## Summary of changes
This test makes the pageserver randomly drop the connection if the
failpoint is enabled, and ensures we can keep querying the pageserver.
This PR also reduces the default timeout to 10 minutes from 3 days.
* lower level on auth success from info to debug (fixes#5820)
* don't log stacktraces on auth errors (as requested on slack). we do this by introducing an `AuthError` type instead of using `anyhow` and `bail`.
* return errors that have been censored for improved security.
## Problem
#5711 and #5367 raced -- the `SlotGuard` type needs `Gate` to properly
enforce its invariant that we may not drop an `Arc<Tenant>` from a slot.
## Summary of changes
Replace the TODO with the intended check of Gate.
## Problem
This test could fail if timing is unlucky, and the deletions in the test
land in two deletion lists instead of one.
## Summary of changes
We await _some_ validations instead of _all_ validations, because our
execution failpoint
will prevent validation proceeding for any but the first DeletionList.
Usually the workload
just generates one, but if it generates two due to timing, then we must
not expect that the
second one will be validated.
neondatabase/autoscaling builds libs/vm-monitor during CI because it's a
necessary component of autoscaling.
workspace_hack includes a lot of crates that are not necessary for
vm-monitor, which artificially inflates the build time on the
autoscaling side, so hopefully removing the dependency should speed
things up.
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
## Problem
- Close#5784
## Summary of changes
- Update the `GetActiveTenantError` -> `QueryError` conversion process
in `pageserver/src/page_service.rs`
- Update the pytest logging exceptions in
`./test_runner/regress/test_tenant_detach.py`
The idea is to pass neon_* prefixed options to control plane. It can be
used by cplane to dynamically create timelines and computes. Such
options also should be excluded from passing to compute. Another issue
is how connection caching is working now, because compute's instance now
depends not only on hostname but probably on such options too I included
them to cache key.
## Problem
For quickly rotating JWT secrets, we want to be able to reload the JWT
public key file in the pageserver, and also support multiple JWT keys.
See #4897.
## Summary of changes
* Allow directories for the `auth_validation_public_key_path` config
param instead of just files. for the safekeepers, all of their config options
also support multiple JWT keys.
* For the pageservers, make the JWT public keys easily globally swappable
by using the `arc-swap` crate.
* Add an endpoint to the pageserver, triggered by a POST to
`/v1/reload_auth_validation_keys`, that reloads the JWT public keys from
the pre-configured path (for security reasons, you cannot upload any
keys yourself).
Fixes#4897
---------
Co-authored-by: Heikki Linnakangas <heikki@neon.tech>
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
## Problem
When enabled, this failpoint would busy-spin in a loop that emits log
messages.
## Summary of changes
Move the failpoint inside a backoff::exponential block: it will still
spam the log, but at much lower rate.
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
## Problem
https://github.com/neondatabase/neon/pull/5576 added `build-tag`
reporting to `libmetrics_build_info`, but it's not reported because we
didn't set the corresponding env variable in the build process.
## Summary of changes
- Add `BUILD_TAG` env var while building services
I forked the project and in my local repo, I wasn't able to compile the
project and in my search, I found the solution in neon forum. After a PR
discussion, I made a change in the makefile to alert the missing `git
submodules update` step.
---------
Signed-off-by: Fernando Luz <prof.fernando.luz@gmail.com>
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
This was preventing it getting cleanly converted to a
CalculateLogicalSizeError::Cancelled, resulting in "Logical size
calculation failed" errors in logs.
## Problem
See: https://github.com/neondatabase/neon/issues/5796
## Summary of changes
Completing the refactor is quite verbose and can be done in stages: each
interface that is currently called directly from a top-level mgr.rs
function can be moved into TenantManager once the relevant subsystems
have access to it.
Landing the initial change to create of TenantManager is useful because
it enables new code to use it without having to be altered later, and
sets us up to incrementally fix the existing code to use an explicit
Arc<TenantManager> instead of relying on the static TENANTS.
## Problem
Some comments in 'receive_wal.rs' is not suitable. It may copy from
'send_wal.rs' and leave it unchanged.
## Summary of changes
This commit fixes two comments in the code:
Changed "/// Unregister walsender." to "/// Unregister walreceiver."
Changed "///Scope guard to access slot in WalSenders registry" to
"///Scope guard to access slot in WalReceivers registry."
Load the metadata from the returned `GetBlobResponse` and avoid
downloading it via a separate request.
As it turns out, the SDK does return the metadata:
https://github.com/Azure/azure-sdk-for-rust/issues/1439 .
This PR will reduce the number of requests to Azure caused by downloads.
Fixes#5571
## Problem
Follows on from #5299
- We didn't have a generic way to protect a tenant undergoing changes:
`Tenant` had states, but for our arbitrary transitions between
secondary/attached, we need a general way to say "reserve this tenant
ID, and don't allow any other ops on it, but don't try and report it as
being in any particular state".
- The TenantsMap structure was behind an async RwLock, but it was never
correct to hold it across await points: that would block any other
changes for all tenants.
## Summary of changes
- Add the `TenantSlot::InProgress` value. This means:
- Incoming administrative operations on the tenant should retry later
- Anything trying to read the live state of the tenant (e.g. a page
service reader) should retry later or block.
- Store TenantsMap in `std::sync::RwLock`
- Provide an extended `get_active_tenant_with_timeout` for page_service
to use, which will wait on InProgress slots as well as non-active
tenants.
Closes: https://github.com/neondatabase/neon/issues/5378
---------
Co-authored-by: Christian Schwarz <christian@neon.tech>
## Problem
When shutting down a Tenant, it isn't just important to cause any
background tasks to stop. It's also important to wait until they have
stopped before declaring shutdown complete, in cases where we may re-use
the tenant's local storage for something else, such as running in
secondary mode, or creating a new tenant with the same ID.
## Summary of changes
A `Gate` class is added, inspired by
[seastar::gate](https://docs.seastar.io/master/classseastar_1_1gate.html).
For types that have an important lifetime that corresponds to some
physical resource, use of a Gate as well as a CancellationToken provides
a robust pattern for async requests & shutdown:
- Requests must always acquire the gate as long as they are using the
object
- Shutdown must set the cancellation token, and then `close()` the gate
to wait for requests in progress before returning.
This is not for memory safety: it's for expressing the difference
between "Arc<Tenant> exists", and "This tenant's files on disk are
eligible to be read/written".
- Both Tenant and Timeline get a Gate & CancellationToken.
- The Timeline gate is held during eviction of layers, and during
page_service requests.
- Existing cancellation support in page_service is refined to use the
timeline-scope cancellation token instead of a process-scope
cancellation token. This replaces the use of `task_mgr::associate_with`:
tasks no longer change their tenant/timelineidentity after being
spawned.
The Tenant's Gate is not yet used, but will be important for
Tenant-scoped operations in secondary mode, where we must ensure that
our secondary-mode downloads for a tenant are gated wrt the activity of
an attached Tenant.
This is part of a broader move away from using the global-state driven
`task_mgr` shutdown tokens:
- less global state where we rely on implicit knowledge of what task a
given function is running in, and more explicit references to the
cancellation token that a particular function/type will respect, making
shutdown easier to reason about.
- eventually avoid the big global TASKS mutex.
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
Improve the serde impl for several types (`Lsn`, `TenantId`,
`TimelineId`) by making them sensitive to
`Serializer::is_human_readadable` (true for json, false for bincode).
Fixes#3511 by:
- Implement the custom serde for `Lsn`
- Implement the custom serde for `Id`
- Add the helper module `serde_as_u64` in `libs/utils/src/lsn.rs`
- Remove the unnecessary attr `#[serde_as(as = "DisplayFromStr")]` in
all possible structs
Additionally some safekeeper types gained serde tests.
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
## Problem
The scrubber didn't know how to find the latest index_part when
generations were in use.
## Summary of changes
- Teach the scrubber to do the same dance that pageserver does when
finding the latest index_part.json
- Teach the scrubber how to understand layer files with generation
suffixes.
- General improvement to testability: scan_metadata has a machine
readable output that the testing `S3Scrubber` wrapper can read.
- Existing test coverage of scrubber was false-passing because it just
didn't see any data due to prefixing of data in the bucket. Fix that.
This is incremental improvement: the more confidence we can have in the
scrubber, the more we can use it in integration tests to validate the
state of remote storage.
---------
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
## Problem
See https://neondb.slack.com/archives/C04DGM6SMTM/p1698226491736459
## Summary of changes
Update WAL affected buffers when restoring WAL from safekeeper
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: Arseny Sher <sher-ars@yandex.ru>
Some of the log messages were lost with the #4938. This PR adds some of
them back, most notably:
- starting to on-demand download
- successful completion of on-demand download
- ability to see when there were many waiters for the layer download
- "unexpectedly on-demand downloading ..." is now `info!`
Additionally some rare events are logged as error, which should never
happen.
## Problem
Proxy doesn't accept wake_compute responses with the allowed IPs.
## Summary of changes
Extend wake_compute api to be able to return allowed_ips.
when introducing `get_and_upgrade` I forgot that an `evict_and_wait`
would had already incremented the counter for started evictions, but an
upgrade would just "silently" cancel the eviction as no drop would ever
run. these metrics are likely sources for alerts with the next release,
so it's important to keep them correct.
In an earlier PR
https://github.com/neondatabase/neon/pull/5743#discussion_r1378625244 I
added a FIXME and there's a simple solution suggested by @jcsp, so
implement it. Wondering why I did not implement this originally, there
is no concept of a permanent failure, so this failure will happen quite
often. I don't think the frequency is a problem however.
Sadly for std::fs::FileType there is only decimal and hex formatting, no
octal.
Following from discussion on
https://github.com/neondatabase/neon/pull/5436 where hacking an implicit
die-on-fatal-io behavior into an Error type was a source of disagreement
-- in this PR, dying on fatal I/O errors is explicit, with `fatal_err`
and `maybe_fatal_err` helpers in the `MaybeFatalIo` trait, which is
implemented for std::io::Result.
To enable this approach with `crashsafe_overwrite`, the return type of
that function is changed to std::io::Result -- the previous error enum
for this function was not used for any logic, and the utility of saying
exactly which step in the function failed is outweighed by the hygiene
of having an I/O funciton return an io::Result.
The initial use case for these helpers is the deletion queue.
With the layer implementation as was done in #4938, it is possible via
cancellation to cause two concurrent downloads on the same path, due to
how `RemoteTimelineClient::download_remote_layer` does tempfiles. Thread
the init semaphore through the spawned task of downloading to make this
impossible to happen.
Right before merging, I added a loop to `fn
LayerInner::get_or_maybe_download`, which was always supposed to be
there. However I had forgotten to restart initialization instead of
waiting for the eviction to happen to support original design goal of
"eviction should always lose to redownload (or init)". This was wrong.
After this fix, if `spawn_blocking` queue is blocked on something,
nothing bad will happen.
Part of #5737.
## Problem
Some requests with `Authorization` header did not properly set the
`Bearer ` prefix. Problem explained here
https://github.com/neondatabase/cloud/issues/6390.
## Summary of changes
Added `Bearer ` prefix to missing requests.
## Problem
test_stderr hangs on MacOS.
See https://neondb.slack.com/archives/C036U0GRMRB/p1698438997903919
## Summary of changes
Always handle POLLHUP to prevent infinite loop.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
The `LayerInner::version` never needed to be read in more than one
place. Clarified while fixing #5737 of which this is the first step.
This decrements possible wrong atomics usage in Layer, but does not
really fix anything.
Minor changes from while I have been working on HA tests:
- Manual pytest executions came with some warnings from `log.warn()`
usage
- When something fails in a generations-enabled test, it it useful to
have a log from the attachment service of what attached when, and with
which generation.
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
Prepare for a new release workflow
* Release PR is created on Fridays
* The discussion/approval happens during Friday
* Sunday morning the deployment will be done in central-il and perf
tests will be run
* On Monday early IST morning gradually start rolling (starting from US
regions as they are still in weekend time)
See slack for discussion:
https://neondb.slack.com/archives/C04P81J55LK/p1698565305607839?thread_ts=1698428241.031979&cid=C04P81J55LK
## Problem
In #5658 we suppressed the first-iteration output from these logs, but
the volume of warnings is still problematic.
## Summary of changes
- Downgrade all slow task warnings to INFO. The information is still
there if we actively want to know about which tasks are running slowly,
without polluting the overall stream of warnings with situations that
are unsurprising to us.
- Revert the previous change so that we output on the first iteration as
we used to do. There is no reason to suppress these, now that the
severity is just info.
## Problem
Neon doesn't compile on nightly and had numerous clippy complaints.
## Summary of changes
1. Fixed troublesome dependency
2. Fixed or ignored the lints where appropriate
## Problem
Role need to have REPLICATION privilege to be able to used for logical
replication.
New roles are created with this option.
This PR tries to update existed roles.
## Summary of changes
Update roles in `handle_roles` method
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
- include Layer generation in the default display, with
Generation::Broken as `-broken`
- omit layer from `layer_gc` span because the api it works with needs to
support N layers, so the api needs to log each layer
## Problem
If there were stray files in the timelines/ dir after tenant deletion,
pageserver could panic on out of range.
## Summary of changes
Use iterator `take()`, which doesn't care if the number of elements
available is less than requested.
## Problem
See
https://neondb.slack.com/archives/C036U0GRMRB/p1698652221399419?thread_ts=1698438997.903919&cid=C036U0GRMRB
## Summary of changes
Check if record pointer is not NULL before trying to print record
descriptor
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
The flush task logs a backtrace if it tries to upload and remote
timeline client is already in stopped state.
Therefore we cannot shut them down concurrently: flush task must be shut
down first.
This wasn't more obvious because:
- Timeline deletions IRL usually happen when not much is being written
- In tests, there is a global allow-list for this log
It's not obvious whether removing the global log allow list is safe,
this PR was prompted by how the log spam got in my way when testing
deletion changes.
## Problem
accidental spam
## Summary of changes
don't spam control plane if control plane is down :)
## Checklist before requesting a review
- [x] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
## Problem
In #5539, I moved the connect_to_compute latency to start counting
before authentication - this is because authentication will perform some
calls to the control plane in order to get credentials and to eagerly
wake a compute server. It felt important to include these times in the
latency metric as these are times we should definitely care about
reducing.
What is not interesting to record in this metric is the roundtrip time
during authentication when we wait for the client to respond.
## Summary of changes
Implement a mechanism to pause the latency timer, resuming on drop of
the pause struct. We pause the timer right before we send the
authentication message to the client, and we resume the timer right
after we complete the authentication flow.
## Problem
## Summary of changes
## Checklist before requesting a review
- [x] I have performed a self-review of my code.
- [ ] ~~If it is a core feature, I have added thorough tests.~~
- [ ] ~~Do we need to implement analytics? if so did you add the
relevant metrics to the dashboard?~~
- [ ] ~~If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.~~
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
#5649 added the concept of dangling layers which #4938 uses but only
partially. I forgot to change `schedule_compaction_update` to not
schedule deletions to uphold the "have a layer, you can read it".
With the now remembered fix, I don't think these checks should ever fail
except for a mistake I already did. These changes might be useful for
protecting future changes, even though the Layer carrying the generation
AND the `schedule_(gc|compaction)_update` require strong arcs.
Rationale for keeping the `#[cfg(feature = "testing")]` is worsening any
leak situation which might come up.
- Add a new util `project_build_tag` macro, similar to
`project_git_version`
- Update the `set_build_info_metric` to accept and make use of
`build_tag` info
- Update all codes which use the `set_build_info_metric`
This benchmark started failing after #5580 merged.
It was manually deleting some local content on a pageserver, and
expecting the behavior that the pageserver would "forget" about the
timeline on startup as a result. That is no longer our behavior:
pageservers use the remote storage as the source of truth.
Rather than having the test go manually delete things at all, we can
just delete the whole tenant via the pageserver API, and thereby start
from a clean situation.
## Problem
The previous garbage cleanup functionality relied on doing a dry run,
inspecting logs, and then doing a deletion. This isn't ideal, because
what one actually deletes might not be the same as what one saw in the
dry run. It's also risky UX to rely on presence/absence of one CLI flag
to control deletion: ideally the deletion command should be totally
separate from the one that scans the bucket.
Related: https://github.com/neondatabase/neon/issues/5037
## Summary of changes
This is a major re-work of the code, which results in a net decrease in
line count of about 600. The old code for removing garbage was build
around the idea of doing discovery and purging together: a
"delete_batch_producer" sent batches into a deleter. The new code writes
out both procedures separately, in functions that use the async streams
introduced in https://github.com/neondatabase/neon/pull/5176 to achieve
fast concurrent access to S3 while retaining the readability of a single
function.
- Add `find-garbage`, which writes out a JSON file of tenants/timelines
to purge
- Add `purge-garbage` which consumes the garbage JSON file, applies some
extra validations, and does deletions.
- The purge command will refuse to execute if the garbage file indicates
that only garbage was found: this guards against classes of bugs where
the scrubber might incorrectly deem everything garbage.
- The purge command defaults to only deleting tenants that were found in
"deleted" state in the control plane. This guards against the risk that
using the wrong console API endpoint could cause all tenants to appear
to be missing.
Outstanding work for a future PR:
- Make whatever changes are needed to adapt to the Console/Control Plane
separation.
- Make purge even safer by checking S3 `Modified` times for
index_part.json files (not doing this here, because it will depend on
the generation-aware changes for finding index_part.json files)
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
Co-authored-by: Shany Pozin <shany@neon.tech>
Only applicable change was neondatabase/autoscaling#584, setting
pgbouncer auth_dbname=postgres in order to fix superuser connections
from preventing dropping databases.
## Problem
We currently require full restart of compute if we change the pageserver
url
## Summary of changes
Makes it so that we don't have to do a full restart, but can just send
SIGHUP
## Problem
The pageserver had two ways of loading a tenant:
- `spawn_load` would trust on-disk content to reflect all existing
timelines
- `spawn_attach` would list timelines in remote storage.
It was incorrect for `spawn_load` to trust local disk content, because
it doesn't know if the tenant might have been attached and written
somewhere else. To make this correct would requires some generation
number checks, but the payoff is to avoid one S3 op per tenant at
startup, so it's not worth the complexity -- it is much simpler to have
one way to load a tenant.
## Summary of changes
- `Tenant` objects are always created with `Tenant::spawn`: there is no
more distinction between "load" and "attach".
- The ability to run without remote storage (for `neon_local`) is
preserved by adding a branch inside `attach` that uses a fallback
`load_local` if no remote_storage is present.
- Fix attaching a tenant when it has a timeline with no IndexPart: this
can occur if a newly created timeline manages to upload a layer before
it has uploaded an index.
- The attach marker file that used to indicate whether a tenant should
be "loaded" or "attached" is no longer needed, and is removed.
- The GenericRemoteStorage interface gets a `list()` method that maps
more directly to what ListObjects does, returning both keys and common
prefixes. The existing `list_files` and `list_prefixes` methods are just
calls into `list()` now -- these can be removed later if we would like
to shrink the interface a bit.
- The remote deletion marker is moved into `timelines/` and detected as
part of listing timelines rather than as a separate GET request. If any
existing tenants have a marker in the old location (unlikely, only
happens if something crashes mid-delete), then they will rely on the
control plane retrying to complete their deletion.
- Revise S3 calls for timeline listing and tenant load to take a
cancellation token, and retry forever: it never makes sense to make a
Tenant broken because of a transient S3 issue.
## Breaking changes
- The remote deletion marker is moved from `deleted` to
`timelines/deleted` within the tenant prefix. Markers in the old
location will be ignored: it is the control plane's responsibility to
retry deletions until they succeed. Markers in the new location will be
tolerated by the previous release of pageserver via
https://github.com/neondatabase/neon/pull/5632
- The local `attaching` marker file is no longer written. Therefore, if
the pageserver is downgraded after running this code, the old pageserver
will not be able to distinguish between partially attached tenants and
fully attached tenants. This would only impact tenants that were partway
through attaching at the moment of downgrade. In the unlikely even t
that we do experience an incident that prompts us to roll back, then we
may check for attach operations in flight, and manually insert
`attaching` marker files as needed.
---------
Co-authored-by: Christian Schwarz <christian@neon.tech>
Previously, if walredo process crashed we would try to spawn a fresh one
every 2 seconds, which is expensive in itself, but also results in a
high I/O load from the part of the compaction prior to the failure,
which we re-run every 2 seconds.
Closes: https://github.com/neondatabase/neon/issues/5671
This fixes issues in pageserver's walredo process where WALRedo
logs of loglevel=LOG are interpreted as errors.
## Problem
See #5560
## Summary of changes
Set the log level to something that doesn't include LOG.
## Problem
Tenant deletions would sometimes be accompanied by compaction stack
traces, because `shutdown()` puts the tenant into stopping state before
it joins background tasks.
## Summary of changes
Treat GC+Compaction as no-ops on a Stopping tenant.
Implement a new `struct Layer` abstraction which manages downloadness
internally, requiring no LayerMap locking or rewriting to download or
evict providing a property "you have a layer, you can read it". The new
`struct Layer` provides ability to keep the file resident via a RAII
structure for new layers which still need to be uploaded. Previous
solution solved this `RemoteTimelineClient::wait_completion` which lead
to bugs like #5639. Evicting or the final local deletion after garbage
collection is done using Arc'd value `Drop`.
With a single `struct Layer` the closed open ended `trait Layer`, `trait
PersistentLayer` and `struct RemoteLayer` are removed following noting
that compaction could be simplified by simply not using any of the
traits in between: #4839.
The new `struct Layer` is a preliminary to remove
`Timeline::layer_removal_cs` documented in #4745.
Preliminaries: #4936, #4937, #5013, #5014, #5022, #5033, #5044, #5058,
#5059, #5061, #5074, #5103, epic #5172, #5645, #5649. Related split off:
#5057, #5134.
## Problem
This line caused lots of errors to be emitted for healthy tenants.
## Summary of changes
Downgrade to debug, since it is an expected code path we'll take for
tenants at startup.
## Problem
We'll need to switch `REMOTE_STORAGE_AZURE_REGION` from the current
`eastus2` region to something `eu-central-1`-like. This may require
changing `AZURE_STORAGE_ACCESS_KEY`.
To make it possible to switch from one place (not to break a lot of
builds on CI), move `REMOTE_STORAGE_AZURE_CONTAINER` and
`REMOTE_STORAGE_AZURE_REGION` to GitHub Variables.
See https://github.com/neondatabase/neon/settings/variables/actions
## Summary of changes
- Get values for `REMOTE_STORAGE_AZURE_CONTAINER` &
`REMOTE_STORAGE_AZURE_REGION` from GitHub Variables
a single operation instead of N uploads and 1 deletion scheduling with
write(layer_map) lock releasing in the between. Compaction update will
make for a much better place to change how the operation will change in
future compared to more general file based operations.
builds upon #5645. solves the problem of difficult to see hopeful
correctness w.r.t. other `index_part.json` changing operations.
Co-authored-by: Shany Pozin <shany@neon.tech>
All loading (attached, or from disk) timelines overwrite the global
gauge for physical size. The `_set` method cannot be used safely, so
remove it and just "add" the physical size.
## Problem
Logical replication requires new AUX_FILES_KEY which is definitely
absent in existed database.
We do not have function to check if key exists in our KV storage.
So I have to handle the error in `list_aux_files` method.
But this key is also included in key space range and accessed y
`create_image_layer` method.
## Summary of changes
Check if AUX_FILES_KEY exists before including it in keyspace.
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: Shany Pozin <shany@neon.tech>
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
This does two things: first a minor refactor to not use HTTP/1.x
style header names and also to not panic if some certain requests had no
"Accept" header. As a second thing, it addresses the third bullet point
from #3689:
> Change `get_lsn_by_timestamp` API method to return LSN even if we only
found commit before the specified timestamp.
This is done by adding a version parameter to the `get_lsn_by_timestamp`
API call and making its behaviour depend on the version number.
Part of #3414 (but doesn't address it in its entirety).
---------
Co-authored-by: Joonas Koivunen <joonas@neon.tech>
## Problem
Our serverless backend was a bit jumbled. As a comment indicated, we
were handling SQL-over-HTTP in our `websocket.rs` file.
I've extracted out the `sql_over_http` and `websocket` files from the
`http` module and put them into a new module called `serverless`.
## Summary of changes
```sh
mkdir proxy/src/serverless
mv proxy/src/http/{conn_pool,sql_over_http,websocket}.rs proxy/src/serverless/
mv proxy/src/http/server.rs proxy/src/http/health_server.rs
mv proxy/src/metrics proxy/src/usage_metrics.rs
```
I have also extracted the hyper server and handler from websocket.rs
into `serverless.rs`
Two of the most common spurious log messages:
- broker connections terminate & we log at error severity. Unfortunately
tonic gives us an "Unknown" error so to suppress these we're doing
string matching. It's hacky but worthwhile for operations.
- the first iteration of tenant background tasks tends to over-run its
schedule and emit a warning. Ultimately we should fix these to run on
time, but for now we are not benefiting from polluting our logs with the
warnings.
## Problem
We started to store test results in a new format in
https://github.com/neondatabase/neon/pull/4549.
This PR switches scripts to query this db.
(we can completely remove old DB/ingestions scripts in a couple of
weeks after the PR merged)
## Summary of changes
- `scripts/benchmark_durations.py` query new database
- `scripts/flaky_tests.py` query new database
## Problem
We defer the returning of connections the the connection pool. It's
possible for our test to be faster than the returning of connections -
which then gets a differing process ID because it opens a new
connection.
## Summary of changes
1. Delay the tests just a little (20ms) to give more chance for
connections to return.
2. Correlate connection IDs with the connection logs a bit more
Quest: #4745. Prerequisite for #4938. Original
https://github.com/neondatabase/neon/pull/4938#issuecomment-1777150665.
The new Layer implementation has so far been using
`RemoteTimelineClient::schedule_layer_file_deletion` from `Layer::drop`
but it was noticed that this could mean that the L0s compaction wanted
to remove could linger in the index part for longer time or be left
there for longer time. Solution is to split the
`RemoteTimelineClient::schedule_layer_file_deletion` into two parts:
- unlinking from index_part.json, to be called from end of compaction
and gc
- scheduling of actual deletions, to be called from `Layer::drop`
The added methods are added unused.
- finally add an `#[instrument]` to Timeline::create_image_layers,
making it easier to see that something is happening because we create
image layers
- format some macro context code
- add a warning not to create new validation functions a la parse do not
validate
Split off from #5198.
Only applicable change was neondatabase/autoscaling#571, removing the
postgres_exporter flags `--auto-discover-databases` and
`--exclude-databases=...`
## Problem
Compaction's source of truth for what layers exist is the LayerManager.
`flush_frozen_layer` updates LayerManager before it has scheduled upload
of the frozen layer.
Compaction can then "see" the new layer, decide to delete it, schedule
uploads of replacement layers, all before `flush_frozen_layer` wakes up
again and schedules the upload. When the upload is scheduled, the local
layer file may be gone, in which case we end up with no such layer in
remote storage, but an entry still added to IndexPart pointing to the
missing layer.
## Summary of changes
Schedule layer uploads inside the `self.layers` lock, so that whenever a
frozen layer is present in LayerManager, it is also present in
RemoteTimelineClient's metadata.
Closes: #5635
## Problem
Looking at logs, I saw more retries being performed for other quota
exceeded errors
## Summary of changes
Filter out all quota exceeded family of errors
The `azure_upload_download_works` test is not cleaning up after itself,
leaving behind the files it is uploading. I found these files when
looking at the contents of the bucket in #5627.
We now clean up the file we uploaded before, like the other tests do it
as well.
Follow-up of #5546
## Problem
https://github.com/neondatabase/neon/pull/5580 will move the remote
deletion marker into the `timelines/` path.
This would cause old pageserver code to fail loading the tenant due to
an apparently invalid timeline ID. That would be a problem if we had to
roll back after deploying #5580
## Summary of changes
If a `deleted` file is in `timelines/` just ignore it.
## Problem
When the number of tenants is large, sequentially issuing the open/read
calls for their config files is a ~1000ms delay during startup. It's not
a lot, but it's simple to fix.
## Summary of changes
Put all the config loads into spawn_blocking() tasks and run them in a
JoinSet. We can simplify this a bit later when we have full async disk
I/O.
---------
Co-authored-by: Shany Pozin <shany@neon.tech>
Before this PR, the ticker was running at default miss behavior `Delay`.
For example, here is the startup output with 25k tenants:
```
2023-10-19T09:57:21.682466Z INFO synthetic_size_worker: starting calculate_synthetic_size_worker
2023-10-19T10:50:44.678202Z WARN synthetic_size_worker: task iteration took longer than the configured period elapsed=3202.995707156s period=10m task=ConsumptionMetricsSyntheticSizeWorker
2023-10-19T10:52:17.408056Z WARN synthetic_size_worker: task iteration took longer than the configured period elapsed=2695.72556035s period=10m task=ConsumptionMetricsSyntheticSizeWorker
```
The first message's `elapsed` value is correct. It matches the
delta between the log line timestamps.
The second one is logged ca 1.5min after, though, but reports a much
larger
`elapsed` than 1.5min.
This PR fixes the behavior by copying what `eviction_task.rs` does.
## Problem
I recently updated the latency timers to include cache miss and pool
miss, as well as connection protocol. By moving the latency timer to
start before authentication, we count a lot more failures and it's
messed up the latency dashboard.
## Summary of changes
Add another label to LatencyTimer metrics for outcome. Explicitly report
on success
## Problem
Every database request locks this cancel map rwlock. At high requests
per second this would have high contention
## Summary of changes
Switch to dashmap which has a sharded rwlock to reduce contention
## Problem
We now persist tenant configuration every time we spawn a tenant. The
persist_tenant_config function is doing a series of non-async filesystem
I/O, because `crashsafe::` isn't async yet. This isn't a demonstrated
problem, but is a source of uncertainty when reasoning about what's
happening with our startup times.
## Summary of changes
- Wrap `crashsafe_overwrite` in `spawn_blocking`.
- Although I think this change makes sense, it does not have a
measurable impact on load time when testing with 10k tenants.
- This can be reverted when we have full async I/O
Stacked atop https://github.com/neondatabase/neon/pull/5559
Before this PR, there was the following race condition:
```
T1: polls for writeable stdin
T1: writes to stdin
T1: enters poll for stdout/stderr
T2: enters poll for stdin write
WALREDO: writes to stderr
KERNEL: wakes up T1 and T2
Tx: reads stderr and prints it
Ty: reads stderr and gets EAGAIN
(valid values for (x, y) are (1, 2) or (2, 1))
```
The concrete symptom that we observed repeatedly was with PG16,
which started logging `registered custom resource manager`
to stderr always, during startup, thereby giving us repeated
opportunity to hit above race condition. PG14 and PG15 didn't log
anything to stderr, hence we could have only hit this race condition
if there was an actual error happening.
This PR fixes the race by moving the reading of stderr into a tokio
task. It exits when the stderr is closed by the child process, which
in turn happens when the child exits, either by itself or because
we killed it.
The downside is that the async scheduling can reorder the log messages,
which can be seen in the new `test_stderr`, which runs in a
single-threaded runtime. I included the output below.
Overall I think we should move the entire walredo to async, as Joonas
proposed many months ago. This PR's asyncification is just the first
step to resolve these
false page reconstruction errors.
After this is fixed, we should stop printing that annoying stderr
message
on walredo startup; it causes noise in the pageserver logs.
That work is tracked in #5399 .
```
2023-10-13T19:05:21.878858Z ERROR apply_wal_records{tenant_id=d546fb76ba529195392fb4d19e243991 pid=753986}: failed to write out the walredo errored input: No such file or directory (os error 2) target=walredo-1697223921878-1132-0.walredo length=1132
2023-10-13T19:05:21.878932Z DEBUG postgres applied 2 WAL records (1062 bytes) in 114666 us to reconstruct page image at LSN 0/0
2023-10-13T19:05:21.878942Z ERROR error applying 2 WAL records 0/16A9388..0/16D4080 (1062 bytes) to base image with LSN 0/0 to reconstruct page image at LSN 0/0 n_attempts=0: apply_wal_records
Caused by:
WAL redo process closed its stdout unexpectedly
2023-10-13T19:05:21.879027Z INFO kill_and_wait_impl{pid=753986}: wait successful exit_status=signal: 11 (SIGSEGV) (core dumped)
2023-10-13T19:05:21.879079Z DEBUG wal-redo-postgres-stderr{pid=753986 tenant_id=d546fb76ba529195392fb4d19e243991 pg_version=16}: wal-redo-postgres stderr_logger_task started
2023-10-13T19:05:21.879104Z ERROR wal-redo-postgres-stderr{pid=753986 tenant_id=d546fb76ba529195392fb4d19e243991 pg_version=16}: received output output="2023-10-13 19:05:21.769 GMT [753986] LOG: registered custom resource manager \"neon\" with ID 134\n"
2023-10-13T19:05:21.879116Z DEBUG wal-redo-postgres-stderr{pid=753986 tenant_id=d546fb76ba529195392fb4d19e243991 pg_version=16}: wal-redo-postgres stderr_logger_task finished
2023-10-13T19:05:22.004439Z ERROR apply_wal_records{tenant_id=d546fb76ba529195392fb4d19e243991 pid=754000}: failed to write out the walredo errored input: No such file or directory (os error 2) target=walredo-1697223922004-1132-0.walredo length=1132
2023-10-13T19:05:22.004493Z DEBUG postgres applied 2 WAL records (1062 bytes) in 125344 us to reconstruct page image at LSN 0/0
2023-10-13T19:05:22.004501Z ERROR error applying 2 WAL records 0/16A9388..0/16D4080 (1062 bytes) to base image with LSN 0/0 to reconstruct page image at LSN 0/0 n_attempts=1: apply_wal_records
Caused by:
WAL redo process closed its stdout unexpectedly
2023-10-13T19:05:22.004588Z INFO kill_and_wait_impl{pid=754000}: wait successful exit_status=signal: 11 (SIGSEGV) (core dumped)
2023-10-13T19:05:22.004624Z DEBUG wal-redo-postgres-stderr{pid=754000 tenant_id=d546fb76ba529195392fb4d19e243991 pg_version=16}: wal-redo-postgres stderr_logger_task started
2023-10-13T19:05:22.004653Z ERROR wal-redo-postgres-stderr{pid=754000 tenant_id=d546fb76ba529195392fb4d19e243991 pg_version=16}: received output output="2023-10-13 19:05:21.884 GMT [754000] LOG: registered custom resource manager \"neon\" with ID 134\n"
2023-10-13T19:05:22.004666Z DEBUG wal-redo-postgres-stderr{pid=754000 tenant_id=d546fb76ba529195392fb4d19e243991 pg_version=16}: wal-redo-postgres stderr_logger_task finished
```
## Problem
## Summary of changes
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
## Problem
Loading tenants shouldn't hang. However, if it does, we shouldn't let
one hung tenant prevent the entire process from starting background
jobs.
## Summary of changes
Generalize the timeout mechanism that we already applied to loading
initial logical sizes: each phase in startup where we wait for a barrier
is subject to a timeout, and startup will proceed if it doesn't complete
within timeout.
Startup metrics will still reflect the time when a phase actually
completed, rather than when we skipped it.
The code isn't the most beautiful, but that kind of reflects the
awkwardness of await'ing on a future and then stashing it to await again
later if we time out. I could imagine making this cleaner in future by
waiting on a structure that doesn't self-destruct on wait() the way
Barrier does, then make InitializationOrder into a structure that
manages the series of waits etc.
Implements fetching of WAL by safekeeper from another safekeeper by imitating
behaviour of last elected leader. This allows to avoid WAL accumulation on
compute and facilitates faster compute startup as it doesn't need to download
any WAL. Actually removing WAL download in walproposer is a matter of another
patch though.
There is a per timeline task which always runs, checking regularly if it should
start recovery frome someone, meaning there is something to fetch and there is
no streaming compute. It then proceeds with fetching, finishing when there is
nothing more to receive.
Implements https://github.com/neondatabase/neon/pull/4875
Fixes an issue we observed on staging that happens when the
autoscaler-agent attempts to immediately downscale the VM after binding,
which is typical for pooled computes.
The issue was occurring because the autoscaler-agent was requesting
downscaling before the vm-monitor had gathered sufficient cgroup memory
stats to be confident in approving it. When the vm-monitor returned an
internal error instead of denying downscaling, the autoscaler-agent
retried the connection and immediately hit the same issue (in part
because cgroup stats are collected per-connection, rather than
globally).
## Problem
Fix elog format error in wallog_mapping_file
## Summary of changes
Use proper case to avoid compilation warning=error in C at MacOS.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
There's currently an issue with the vm-monitor on staging that's not
really feasible to debug because the current display impl gives no
context to the errors (just says "failed to downscale").
Logging the full error should help.
For communications with the autoscaler-agent, it's ok to only provide
the outermost cause, because we can cross-reference with the VM logs.
At some point in the future, we may want to change that.
Create Rust bindings for C functions from walproposer. This allows to
write better tests with real walproposer code without spawning multiple
processes and starting up the whole environment.
`make walproposer-lib` stage was added to build static libraries
`libwalproposer.a`, `libpgport.a`, `libpgcommon.a`. These libraries can
be statically linked to any executable to call walproposer functions.
`libs/walproposer/src/walproposer.rs` contains
`test_simple_sync_safekeepers` to test that walproposer can be called
from Rust to emulate sync_safekeepers logic. It can also be used as a
usage example.
There is a bunch of duplication and manual Result handling that can be
simplified by moving the error conversion into a shared function, using
`map_err`, and the question mark operator.
## Problem
See #5468.
## Summary of changes
Add a new `get_timestamp_of_lsn` endpoint, returning the timestamp
associated with the given lsn.
Fixes#5468.
---------
Co-authored-by: Shany Pozin <shany@neon.tech>
## Problem
S3 can give us a 500 whenever it likes: when this happens at request
level we eat it in `backoff::retry`, but when it happens for a key
inside a DeleteObjects request, we log it at warn level.
## Summary of changes
Allow-list this class of log message in all tests.
## Problem
See https://github.com/neondatabase/company_projects/issues/111
## Summary of changes
Save logical replication files in WAL at compute and include them in
basebackup at pate server.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: Arseny Sher <sher-ars@yandex.ru>
## Problem
- QueryError always logged at error severity, even though disconnections
are not true errors.
- QueryError type is not expressive enough to distinguish actual errors
from shutdowns.
- In some functions we're returning Ok(()) on shutdown, in others we're
returning an error
## Summary of changes
- Add QueryError::Shutdown and use it in places we check for
cancellation
- Adopt consistent Result behavior: disconnects and shutdowns are always
QueryError, not ok
- Transform shutdown+disconnect errors to Ok(()) at the very top of the
task that runs query handler
- Use the postgres protocol error code for "admin shutdown" in responses
to clients when we are shutting down.
Closes: #5517
tl;dr it's really hard to avoid throttling from memory.high, and it
counts tmpfs & page cache usage, so it's also hard to make sense of.
In the interest of fixing things quickly with something that should be
*good enough*, this PR switches to instead periodically fetch memory
statistics from the cgroup's memory.stat and use that data to determine
if and when we should upscale.
This PR fixes#5444, which has a lot more detail on the difficulties
we've hit with memory.high. This PR also supersedes #5488.
## Problem
We need a flag to require proxy protocol (prerequisite for #5416)
## Summary of changes
Add a cli flag to require client IP addresses. Error if IP address is
missing when the flag is active.
## Problem
Transactions break connections in the pool
fixes#4698
## Summary of changes
* Pool `Client`s are smart object that return themselves to the pool
* Pool `Client`s can be 'discard'ed
* Pool `Client`s are discarded when certain errors are encountered.
* Pool `Client`s are discarded when ReadyForQuery returns a non-idle
state.
This test is listing files in a timeline and then evicting them: if the
test ran slowly this could encounter temp files for unfinished
downloads: fix by filtering these out in evict_all_layers.
Before this PR, when we restarted pageserver, we'd see a rush of
`$number_of_tenants` concurrent eviction tasks starting to do imitate
accesses building up in the period of `[init_order allows activations,
$random_access_delay + EvictionPolicyLayerAccessThreshold::period]`.
We simply cannot handle that degree of concurrent IO.
We already solved the problem for compactions by adding a semaphore.
So, this PR shares that semaphore for use by evictions.
Part of https://github.com/neondatabase/neon/issues/5479
Which is again part of https://github.com/neondatabase/neon/issues/4743
Risks / Changes In System Behavior
==================================
* we don't do evictions as timely as we currently do
* we log a bunch of warnings about eviction taking too long
* imitate accesses and compactions compete for the same concurrency
limit, so, they'll slow each other down through this shares semaphore
Changes
=======
- Move the `CONCURRENT_COMPACTIONS` semaphore into `tasks.rs`
- Rename it to `CONCURRENT_BACKGROUND_TASKS`
- Use it also for the eviction imitate accesses:
- Imitate acceses are both per-TIMELINE and per-TENANT
- The per-TENANT is done through coalescing all the per-TIMELINE
tasks via a tokio mutex `eviction_task_tenant_state`.
- We acquire the CONCURRENT_BACKGROUND_TASKS permit early, at the
beginning of the eviction iteration, much before the imitate
acesses start (and they may not even start at all in the given
iteration, as they happen only every $threshold).
- Acquiring early is **sub-optimal** because when the per-timline
tasks coalesce on the `eviction_task_tenant_state` mutex,
they are already holding a CONCURRENT_BACKGROUND_TASKS permit.
- It's also unfair because tenants with many timelines win
the CONCURRENT_BACKGROUND_TASKS more often.
- I don't think there's another way though, without refactoring
more of the imitate accesses logic, e.g, making it all per-tenant.
- Add metrics for queue depth behind the semaphore.
I found these very useful to understand what work is queued in the
system.
- The metrics are tagged by the new `BackgroundLoopKind`.
- On a green slate, I would have used `TaskKind`, but we already had
pre-existing labels whose names didn't map exactly to task kind.
Also the task kind is kind of a lower-level detail, so, I think
it's fine to have a separate enum to identify background work kinds.
Future Work
===========
I guess I could move the eviction tasks from a ticker to "sleep for
$period".
The benefit would be that the semaphore automatically "smears" the
eviction task scheduling over time, so, we only have the rush on restart
but a smeared-out rush afterward.
The downside is that this perverts the meaning of "$period", as we'd
actually not run the eviction at a fixed period. It also means the the
"took to long" warning & metric becomes meaningless.
Then again, that is already the case for the compaction and gc tasks,
which do sleep for `$period` instead of using a ticker.
Fixes#5172 as it:
- removes recoinciliation with remote index_part.json and accepts remote
index_part.json as the truth, deleting any local progress which is yet
to be reflected in remote
- moves to prefer remote metadata
Additionally:
- tests with single LOCAL_FS parametrization are cleaned up
- adds a test case for branched (non-bootstrap) local only timeline
availability after restart
---------
Co-authored-by: Christian Schwarz <christian@neon.tech>
Co-authored-by: John Spray <john@neon.tech>
## Problem
I found some issues with the `/location_config` API when writing new
tests.
## Summary of changes
- Calling the API with the "Detached" state is now idempotent.
- `Tenant::spawn_attach` now takes a boolean to indicate whether to
expect a marker file. Marker files are used in the old attach path, but
not in the new location conf API. They aren't needed because in the New
World, the choice of whether to attach via remote state ("attach") or to
trust local state ("load") will be revised to cope with the transitions
between secondary & attached (see
https://github.com/neondatabase/neon/issues/5550). It is okay to merge
this change ahead of that ticket, because the API is not used in the
wild yet.
- Instead of using `schedule_local_tenant_processing`, the location conf
API handler does its own directory creation and calls `spawn_attach`
directly.
- A new `unsafe_create_dir_all` is added. This differs from
crashsafe::create_dir_all in two ways:
- It is intentionally not crashsafe, because in the location conf API we
are no longer using directory or config existence as the signal for any
important business logic.
- It is async and uses `tokio::fs`.
Before this PR, the following race condition existed:
```
T1: does the apply_wal_records() call and gets back an error
T2: does the apply_wal_records() call and gets back an error
T2: does the kill_and_shutdown
T2: new loop iteration
T2: launches new walredo process
T1: does the kill_and_shutdown of the new process
```
That last step is wrong, T2 already did the kill_and_shutdown.
The symptom of this race condition was that T2 would observe an error
when it tried to do something with the process after T1 killed it.
For example, but not limited to:
`POLLHUP` / `"WAL redo process closed its stderr unexpectedly"`.
The fix in this PR is the following:
* Use Arc to represent walredo processes.
The Arc lives at least as long as the walredo process.
* Use Arc::ptr_eq to determine whether to kill the process or not.
The price is an additional RwLock to protect the new `redo_process`
field
that holds the Arc. I guess that could perhaps be an atomic pointer
swap some day. But, let's get one race fixed without risking introducing
a new one.
The use of Arc/drop is also not super great here because it now allows
for an unlimited number of to-be-killed processes to exist concurrently.
See the various `NB` comments above `drop(proc)` for why it's "ok" right
now due to the blocking `wait` inside `drop`.
Note: an earlier fix attempt was
https://github.com/neondatabase/neon/pull/5545
where we apply_batch_postgres would compare stdout_fd for equality.
That's incorrect because the kernel can reuse the file descriptor when
T2 launches the new process.
Details:
https://github.com/neondatabase/neon/pull/5545#pullrequestreview-1676589373
## Problem
In 89275f6c1e we fixed an issue, when we were dropping db in Postgres
even though cplane request failed. Yet, it introduced a new problem that
we now de-register db in cplane even if we didn't actually drop it in
Postgres.
## Summary of changes
Here we revert extension change, so we now again may leave db in invalid
state after failed drop. Instead, `compute_ctl` is now responsible for
cleaning up invalid databases during full configuration. Thus, there are
two ways of recovering from failed DROP DATABASE:
1. User can just repeat DROP DATABASE, same as in Vanilla Postgres.
2. If they didn't, then on next full configuration (dbs / roles changes
in the API; password reset; or data availability check) invalid db
will be cleaned up in the Postgres and re-created by `compute_ctl`. So
again it follows pretty much the same semantics as Vanilla Postgres --
you need to drop it again after failed drop.
That way, we have a recovery trajectory for both problems.
See this commit for info about `invalid` db state:
a4b4cc1d60
According to it:
> An invalid database cannot be connected to anymore, but can still be
dropped.
While on it, this commit also fixes another issue, when `compute_ctl`
was trying to connect to databases with `ALLOW CONNECTIONS false`. Now
it will just skip them.
Fixes#5435
## Problem
Currently it's unclear how much of the `initial_tenant_load` period is
in S3 objects, and therefore how impactful it is to make changes to
remote operations during startup.
## Summary of changes
- `Tenant::load` is refactored to load remote indices in parallel and to
wait for all these remote downloads to finish before it proceeds to
construct any `Timeline` objects.
- `pageserver_startup_duration_seconds` gets a new `phase` value of
`initial_tenant_load_remote` which counts the time from startup to when
the last tenant finishes loading remote content.
- `test_pageserver_restart` is extended to validate this phase. The
previous version of the test was relying on order of dict entries, which
stopped working when adding a phase, so this is refactored a bit.
- `test_pageserver_restart` used to explicitly create a branch, now it
uses the default initial_timeline. This avoids startup getting held up
waiting for logical sizes, when one of the branches is not in use.
Adds prototype-level support for [Azure blob storage](https://azure.microsoft.com/en-us/products/storage/blobs). Some corners were cut, see the TODOs and the followup issue #5567 for details.
Steps to try it out:
* Create a storage account with block blobs (this is a per-storage
account setting).
* Create a container inside that storage account.
* Set the appropriate env vars: `AZURE_STORAGE_ACCOUNT,
AZURE_STORAGE_ACCESS_KEY, REMOTE_STORAGE_AZURE_CONTAINER,
REMOTE_STORAGE_AZURE_REGION`
* Set the env var `ENABLE_REAL_AZURE_REMOTE_STORAGE=y` and run `cargo
test -p remote_storage azure`
Fixes #5562
## Problem
Bug was introduced by me in 83ae2bd82c
When eviction constructs a RemoteLayer to replace the layer it just
evicted, it is building a LayerFileMetadata using its _current_
generation, rather than the generation of the layer.
## Summary of changes
- Retrieve Generation from RemoteTimelineClient when evicting. This will
no longer be necessary when #4938 lands.
- Add a test for the scenario in question (this fails without the fix).
For 2 weeks we've seen rare, spurious, not-reproducible page
reconstruction
failures with PG16 in prod.
One of the commits we deployed this week was
Commit
commit fc467941f9
Author: Joonas Koivunen <joonas@neon.tech>
Date: Wed Oct 4 16:19:19 2023 +0300
walredo: log retryed error (#546)
With the logs from that commit, we learned that some read() or write()
system call that walredo does fails with `EAGAIN`, aka
`Resource temporarily unavailable (os error 11)`.
But we have no idea where exactly in the code we get back that error.
So, use anyhow instead of fake std::io::Error's as an easy way to get
a backtrace when the error happens, and change the logging to print
that backtrace (i.e., use `{:?}` instead of
`utils::error::report_compact_sources(e)`).
The `WalRedoError` type had to go because we add additional `.context()`
further up the call chain before we `{:?}`-print it. That additional
`.context()` further up doesn't see that there's already an
anyhow::Error
inside the `WalRedoError::ApplyWalRecords` variant, and hence captures
another backtrace and prints that one on `{:?}`-print instead of the
original one inside `WalRedoError::ApplyWalRecords`.
If we ever switch back to `report_compact_sources`, we should make sure
we have some other way to uniquely identify the places where we return
an error in the error message.
## Problem
Currently proxy doesn't handle array of json parameters correctly.
## Summary of changes
Added one more level of quotes escaping for the array of jsons case.
Resolves: https://github.com/neondatabase/neon/issues/5515
## Problem
The `Deleter` currently doesn't use a backoff::retry because it doesn't
need to: it is already inside a loop when doing the deletion, so can
just let the loop go around.
However, this is a problem for logging, because we log on errors, which
includes things like 503/429 cases that would usually be swallowed by a
backoff::retry in most places we use the RemoteStorage interface.
The underlying problem is that RemoteStorage doesn't have a proper error
type, and an anyhow::Error can't easily be interrogated for its original
S3 SdkError because downcast_ref requires a concrete type, but SdkError
is parametrized on response type.
## Summary of changes
Wrap remote deletions in Deleter in a backoff::retry to avoid logging
warnings on transient 429/503 conditions, and for symmetry with how
RemoteStorage is used in other places.
## Problem
When we got an error in the payload of a DeleteObjects response, we only
logged how many errors, not what they were.
## Summary of changes
Log up to 10 specific errors. We do not log all of them because that
would be up to 1000 log lines per request.
## Problem
These types/functions are public and it prevents clippy from catching
unused things.
## Summary of changes
Move to `pub(crate)` and remove the error enum that becomes clearly
unused as a result.
## Problem
- `test_heavy_write_workload` is flaky, and fails because of to
statement timeout
- `test_wal_lagging` is flaky and fails because of the default pytest
timeout (see https://github.com/neondatabase/neon/issues/5305)
## Summary of changes
- `test_heavy_write_workload`: increase statement timeout to 5 minutes
(from default 2 minutes)
- `test_wal_lagging`: increase pytest timeout to 600s (from default
300s)
## Problem
Hard to tell
1. How many clients are connected to proxy
2. How many requests clients are making
3. How many connections are made to a database
1 and 2 are different because of the properties of HTTP.
We have 2 already tracked through `proxy_accepted_connections_total` and
`proxy_closed_connections_total`, but nothing for 1 and 3
## Summary of changes
Adds 2 new counter gauges.
*
`proxy_opened_client_connections_total`,`proxy_closed_client_connections_total`
- how many client connections are open to proxy
*
`proxy_opened_db_connections_total`,`proxy_closed_db_connections_total`
- how many active connections are made through to a database.
For TCP and Websockets, we expect all 3 of these quantities to be
roughly the same, barring users connecting but with invalid details.
For HTTP:
* client_connections/connections can differ because the client
connections can be reused.
* connections/db_connections can differ because of connection pooling.
## Problem
close https://github.com/neondatabase/neon/issues/4702
## Summary of changes
This PR adds a new metrics for wake up errors and breaks it down by most
common reasons (mostly follows the `could_retry` implementation).
## Problem
If a caller detaches a tenant and then attaches it again, pending
deletions from the old attachment might not have happened yet. This is
not a correctness problem, but it causes:
- Risk of leaking some objects in S3
- Some warnings from the deletion queue when pending LSN updates and
pending deletions don't pass validation.
## Summary of changes
- Deletion queue now uses UnboundedChannel so that the push interfaces
don't have to be async.
- This was pulled out of https://github.com/neondatabase/neon/pull/5397,
where it is also useful to be able to drive the queue from non-async
contexts.
- Why is it okay for this to be unbounded? The only way the
unbounded-ness of the channel can become a problem is if writing out
deletion lists can't keep up, but if the system were that overloaded
then the code generating deletions (GC, compaction) would also be
impacted.
- DeletionQueueClient gets a new `flush_advisory` function, which is
like flush_execute, but doesn't wait for completion: this is appropriate
for use in contexts where we would like to encourage the deletion queue
to flush, but don't need to block on it.
- This function is also expected to be useful in next steps for seamless
migration, where the option to flush to S3 while transitioning into
AttachedStale will also include flushing deletion queue, but we wouldn't
want to block on that flush.
- The tenant_detach code in mgr.rs invokes flush_advisory after stopping
the `Tenant` object.
---------
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
## Problem
We don't want to retry customer quota exhaustion errors.
## Summary of changes
Make sure both types of quota exhaustion errors are not retried
## Problem
See https://neondb.slack.com/archives/C05U648A9NJ
In case of failure of flush in prefetch,
prefetch state is reseted. We need to retry register buffer attempt,
otherwise we will get assertion failure.
## Checklist before requesting a review
- [ ] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
## Checklist before merging
- [ ] Do not forget to reformat commit message to not include the above
checklist
---------
Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
## Problem
Seen in CI for https://github.com/neondatabase/neon/pull/5453 -- the
time gap between validation completing and the header getting written is
long enough to fail the test, where it was doing a cheeky 1 second
sleep.
## Summary of changes
- Replace 1 second sleep with a wait_until to see the header file get
written
- Use enums as test params to make the results more readable (instead of
True-False parameters)
- Fix the temp suffix used for deletion queue headers: this worked fine,
but resulted in `..tmp` extension.
## Problem
Writes to the postgres client socket from the page server were not
wrapped in cancellation handling, so a stuck client connection could
prevent tenant shutdowwn.
## Summary of changes
All the places we call flush() to write to the socket, we should be
respecting the cancellation token for the task.
In this PR, I explicitly pass around a CancellationToken rather than
doing inline `task_mgr::shutdown_token` calls, to avoid coupling it to
the global task_mgr state and make it easier to refactor later.
I have some follow-on commits that add a Shutdown variant to QueryError
and use it more extensively, but that's pure refactor so will keep
separate from this bug fix PR.
Closes: https://github.com/neondatabase/neon/issues/5341
Part of #5172. First commits show that we used to allow starting up a
compute or creating a branch off a not yet uploaded timeline. This PR
moves activation of a timeline to happen **after** initial layer file(s)
(if any) and `index_part.json` have been uploaded. Simply moving
activation to be *after* downloads have finished works because we now
spawn a task per http request handler.
Current behaviour of uploading on the timelines on next startup is kept,
to be removed later as part of #5172.
Adds:
- `NeonCli.map_branch` and corresponding `neon_local` implementation:
allow creating computes for timelines managed via pageserver http
client/api
- possibly duplicate tests (I did not want to search for, will cleanup
in a follow-up if these duplicated)
Changes:
- make `wait_until_tenant_state` return immediatedly on `Broken` and not
wait more
## Problem
Fix neon_local pageserver status command
#5430
## Summary of changes
Fix clap config for pageserver status subcommand
## Checklist before requesting a review
- [x] I have performed a self-review of my code.
- [ ] If it is a core feature, I have added thorough tests.
- [ ] Do we need to implement analytics? if so did you add the relevant
metrics to the dashboard?
- [ ] If this PR requires public announcement, mark it with
/release-notes label and add several sentences in this section.
Signed-off-by: Rahul Modpur <rmodpur2@gmail.com>
2023-10-09 09:13:57 +01:00
545 changed files with 73111 additions and 27076 deletions
# TODO: move deployPreprodRegion to release (`"$GITHUB_REF_NAME" == "release"` block), once Staging support different compute tag prefixes for different regions
gh workflow --repo neondatabase/aws run deploy-dev.yml --ref main -f branch=main -f dockerTag=${{needs.tag.outputs.build-tag}} -f deployPreprodRegion=true
elif [[ "$GITHUB_REF_NAME" == "release" ]]; then
gh workflow --repo neondatabase/aws run deploy-prod.yml --ref main -f branch=main -f dockerTag=${{needs.tag.outputs.build-tag}} -f disclamerAcknowledged=true
gh workflow --repo neondatabase/aws run deploy-prod.yml --ref main -f branch=main -f dockerTag=${{needs.tag.outputs.build-tag}}
else
echo "GITHUB_REF_NAME (value '$GITHUB_REF_NAME') is not set to either 'main' or 'release'"
RUN curl -fsSL "https://github.com/protocolbuffers/protobuf/releases/download/v${PROTOC_VERSION}/protoc-${PROTOC_VERSION}-linux-$(uname -m | sed 's/aarch64/aarch_64/g').zip" -o "protoc.zip"\
@@ -14,8 +14,8 @@ Alternatively, compile and run the project [locally](#running-local-installation
A Neon installation consists of compute nodes and the Neon storage engine. Compute nodes are stateless PostgreSQL nodes backed by the Neon storage engine.
The Neon storage engine consists of two major components:
- Pageserver. Scalable storage backend for the compute nodes.
- Safekeepers. The safekeepers form a redundant WAL service that received WAL from the compute node, and stores it durably until it has been processed by the pageserver and uploaded to cloud storage.
- Pageserver: Scalable storage backend for the compute nodes.
- Safekeepers: The safekeepers form a redundant WAL service that received WAL from the compute node, and stores it durably until it has been processed by the pageserver and uploaded to cloud storage.
See developer documentation in [SUMMARY.md](/docs/SUMMARY.md) for more information.
@@ -29,13 +29,14 @@ See developer documentation in [SUMMARY.md](/docs/SUMMARY.md) for more informati
* On Arch based systems, these packages are needed:
```bash
@@ -80,9 +81,9 @@ The project uses [rust toolchain file](./rust-toolchain.toml) to define the vers
This file is automatically picked up by [`rustup`](https://rust-lang.github.io/rustup/overrides.html#the-toolchain-file) that installs (if absent) and uses the toolchain version pinned in the file.
rustup users who want to build with another toolchain can use [`rustup override`](https://rust-lang.github.io/rustup/overrides.html#directory-overrides) command to set a specific toolchain for the project's directory.
rustup users who want to build with another toolchain can use the [`rustup override`](https://rust-lang.github.io/rustup/overrides.html#directory-overrides) command to set a specific toolchain for the project's directory.
non-rustup users most probably are not getting the same toolchain automatically from the file, so are responsible to manually verify their toolchain matches the version in the file.
non-rustup users most probably are not getting the same toolchain automatically from the file, so are responsible to manually verify that their toolchain matches the version in the file.
Newer rustc versions most probably will work fine, yet older ones might not be supported due to some new features used by the project or the crates.
#### Building on Linux
@@ -123,7 +124,7 @@ make -j`sysctl -n hw.logicalcpu` -s
To run the `psql` client, install the `postgresql-client` package or modify `PATH` and `LD_LIBRARY_PATH` to include `pg_install/bin` and `pg_install/lib`, respectively.
To run the integration tests or Python scripts (not required to use the code), install
Python (3.9 or higher), and install python3 packages using `./scripts/pysync` (requires [poetry>=1.3](https://python-poetry.org/)) in the project directory.
Python (3.9 or higher), and install the python3 packages using `./scripts/pysync` (requires [poetry>=1.3](https://python-poetry.org/)) in the project directory.
#### Running neon database
@@ -149,6 +150,9 @@ tenant 9ef87a5bf0d92544f6fafeeb3239695c successfully created on the pageserver
Created an initial timeline 'de200bd42b49cc1814412c7e592dd6e9' at Lsn 0/16B5A50 for tenant: 9ef87a5bf0d92544f6fafeeb3239695c
Setting tenant 9ef87a5bf0d92544f6fafeeb3239695c as a default one
# create postgres compute node
> cargo neon endpoint create main
# start postgres compute node
> cargo neon endpoint start main
Starting new endpoint main (PostgreSQL v14) on timeline de200bd42b49cc1814412c7e592dd6e9 ...
@@ -162,7 +166,7 @@ Starting postgres at 'postgresql://cloud_admin@127.0.0.1:55432/postgres'
2. Now, it is possible to connect to postgres and run some queries:
bail!("attempting to create a duplicate primary endpoint on tenant {tenant_id}, timeline {timeline_id}: endpoint {key:?} exists already. please don't do this, it is not supported.");
@@ -21,7 +21,7 @@ We build all images after a successful `release` tests run and push automaticall
## Docker Compose example
You can see a [docker compose](https://docs.docker.com/compose/) example to create a neon cluster in [/docker-compose/docker-compose.yml](/docker-compose/docker-compose.yml). It creates the following conatainers.
You can see a [docker compose](https://docs.docker.com/compose/) example to create a neon cluster in [/docker-compose/docker-compose.yml](/docker-compose/docker-compose.yml). It creates the following containers.
- pageserver x 1
- safekeeper x 3
@@ -38,7 +38,7 @@ You can specify version of neon cluster using following environment values.
- TAG: the tag version of [docker image](https://registry.hub.docker.com/r/neondatabase/neon/tags) (default is latest), which is tagged in [CI test](/.github/workflows/build_and_test.yml)
```
$ cd docker-compose/
$ docker-compose down # remove the conainers if exists
$ docker-compose down # remove the containers if exists
$ PG_VERSION=15 TAG=2937 docker-compose up --build -d # You can specify the postgres and image version
Creating network "dockercompose_default" with the default driver
Currently we have `cloud` repository that contains code implementing public API for our clients as well as code for managing storage and internal infrastructure services. We can split everything user-related from everything storage-related to make it easier to test and maintain.
This RFC proposes to introduce a new control-plane service with HTTP API. The overall architecture will look like this:
```markup
. x
external area x internal area
(our clients) x (our services)
x
x ┌───────────────────────┐
x ┌───────────────┐ > ┌─────────────────────┐ │ Storage (EC2) │
x │ console db │ > │ control-plane db │ │ │
x └───────────────┘ > └─────────────────────┘ │ - safekeepers │
x ▲ > ▲ │ - pageservers │
x │ > │ │ │
┌──────────────────┐ x ┌───────┴───────┐ > │ │ Dependencies │
x │ console ├───────►│ control-plane ├────►│ - S3 │
┌──────────────────┐ x │ │ > │ (deployed in k8s) │ │ - more? │
│public API clients├──►│ │ > │ │ │ │
└──────────────────┘ x └───────┬───────┘ > └──────────┬──────────┘ └───────────────────────┘
x │ > ▲ │ ▲
x │ > │ │ │
x ┌───────┴───────┐ > │ │ ┌───────────┴───────────┐
x │ dependencies │ > │ │ │ │
x │- analytics │ > │ └───────────────►│ computes │
x │- auth │ > │ │ (deployed in k8s) │
x │- billing │ > │ │ │
x └───────────────┘ > │ └───────────────────────┘
x > │ ▲
x > ┌─────┴───────────────┐ │
┌──────────────────┐ x > │ │ │
│ │ x > │ proxy ├─────────────────┘
│ postgres ├───────────────────────────►│ (deployed in k8s) │
│ users │ x > │ │
│ │ x > └─────────────────────┘
└──────────────────┘ x >
>
>
closed-source > open-source
>
>
```
Notes:
- diagram is simplified in the less-important places
- directed arrows are strict and mean that connections in the reverse direction are forbidden
This split is quite complex and this RFC proposes several smaller steps to achieve the larger goal:
1. Start by refactoring the console code, the goal is to have console and control-plane code in the different directories without dependencies on each other.
2. Do similar refactoring for tables in the console database, remove queries selecting data from both console and control-plane; move control-plane tables to a separate database.
3. Implement control-plane HTTP API serving on a separate TCP port; make all console→control-plane calls to go through that HTTP API.
4. Move control-plane source code to the neon repo; start control-plane as a separate service.
## Motivation
These are the two most important problems we want to solve:
- Publish open-source implementation of all our cloud/storage features
- Make a unified control-plane that is used in all cloud (serverless) and local (tests) setups
Right now we have some closed-source code in the cloud repo. That code contains implementation for running Neon computes in k8s and without that code it’s impossible to automatically scale PostgreSQL computes. That means that we don’t have an open-source serverless PostgreSQL at the moment.
After splitting and open-sourcing control-plane service we will have source code and Docker images for all storage services. That control-plane service should have HTTP API for creating and managing tenants (including all our storage features), while proxy will listen for incoming connections and create computes on-demand.
Improving our test suite is an important task, but requires a lot of prerequisites and may require a separate RFC. Possible implementation of that is described in the section [Next steps](#next-steps).
Another piece of motivation can be a better involvement of storage development team into a control-plane. By splitting control-plane from the console, it can be more convenient to test and develop control-plane with paying less attention to “business” features, such as user management, billing and analytics.
For example, console currently requires authentication providers such as GitHub OAuth to work at all, as well as nodejs to be able to build it locally. It will be more convenient to build and run it locally without these requirements.
## Proposed implementation
### Current state of things
Let’s start with defining the current state of things at the moment of this proposal. We have three repositories containing source code:
- open-source `postgres` — our fork of postgres
- open-source `neon` — our main repository for storage source code
- closed-source `cloud` — mostly console backend and UI frontend
This proposal aims not to change anything at the existing code in `neon` and `postgres` repositories, but to create control-plane service and move it’s source code from `cloud` to the `neon` repository. That means that we need to split code in `cloud` repo only, and will consider only this repository for exploring its source code.
Let’s look at the miscellaneous things in the `cloud` repo which are NOT part of the console application, i.e. NOT the Go source code that is compiled to the `./console` binary. There we have:
- command-line tools, such as cloudbench, neonadmin
And also let’s take a look at what we have in the console source code, which is the service we’d like to split:
- API Servers
- Public API v2
- Management API v2
- Public API v1
- Admin API v1 (same port as Public API v1)
- Management API v1
- Workers
- Monitor Compute Activity
- Watch Failed Operations
- Availability Checker
- Business Metrics Collector
- Internal Services
- Auth Middleware, UserIsAdmin, Cookies
- Cable Websocket Server
- Admin Services
- Global Settings, Operations, Pageservers, Platforms, Projects, Safekeepers, Users
- Authenticate Proxy
- API Keys
- App Controller, serving UI HTML
- Auth Controller
- Branches
- Projects
- Psql Connect + Passwordless login
- Users
- Cloud Metrics
- User Metrics
- Invites
- Pageserver/Safekeeper management
- Operations, k8s/docker/common logic
- Platforms, Regions
- Project State
- Projects Roles, SCRAM
- Global Settings
- Other things
- segment analytics integration
- sentry integration
- other common utilities packages
### Drawing the splitting line
The most challenging and the most important thing is to define the line that will split new control-plane service from the existing cloud service. If we don’t get it right, then we can end up with having a lot more issues without many benefits.
We propose to define that line as follows:
- everything user-related stays in the console service
- everything storage-related should be in the control-plane service
- something that falls in between should be decided where to go, but most likely should stay in the console service
- some similar parts should be in both services, such as admin/management/db_migrations
We call user-related all requests that can be connected to some user. The general idea is don’t have any user_id in the control-plane service and operate exclusively on tenant_id+timeline_id, the same way as existing storage services work now (compute, safekeeper, pageserver).
Storage-related things can be defined as doing any of the following:
- using k8s API
- doing requests to any of the storage services (proxy, compute, safekeeper, pageserver, etc..)
- tracking current status of tenants/timelines, managing lifetime of computes
Based on that idea, we can say that new control-plane service should have the following components:
- single HTTP API for everything
- Create and manage tenants and timelines
- Manage global settings and storage configuration (regions, platforms, safekeepers, pageservers)
- Admin API for storage health inspection and debugging
- Workers
- Monitor Compute Activity
- Watch Failed Operations
- Availability Checker
- Internal Services
- Admin Services
- Global Settings, Operations, Pageservers, Platforms, Tenants, Safekeepers
- Authenticate Proxy
- Branches
- Psql Connect
- Cloud Metrics
- Pageserver/Safekeeper management
- Operations, k8s/docker/common logic
- Platforms, Regions
- Tenant State
- Compute Roles, SCRAM
- Global Settings
---
And other components should probably stay in the console service:
- API Servers (no changes here)
- Public API v2
- Management API v2
- Public API v1
- Admin API v1 (same port as Public API v1)
- Management API v1
- Workers
- Business Metrics Collector
- Internal Services
- Auth Middleware, UserIsAdmin, Cookies
- Cable Websocket Server
- Admin Services
- Users admin stays the same
- Other admin services can redirect requests to the control-plane
- API Keys
- App Controller, serving UI HTML
- Auth Controller
- Projects
- User Metrics
- Invites
- Users
- Passwordless login
- Other things
- segment analytics integration
- sentry integration
- other common utilities packages
There are also miscellaneous things that are useful for all kinds of services. So we can say that these things can be in both services:
- markdown documentation
- e2e python tests
- make build scripts, code generation scripts
- database migrations
- swagger definitions
The single entrypoint to the storage should be control-plane API. After we define that API, we can have code-generated implementation for the client and for the server. The general idea is to move code implementing storage components from the console to the API implementation inside the new control-plane service.
After the code is moved to the new service, we can fill the created void by making API calls to the new service:
- authorization of the client
- mapping user_id + project_id to the tenant_id
- calling the control-plane API
### control-plane API
Currently we have the following projects API in the console:
```
GET /projects/{project_id}
PATCH /projects/{project_id}
POST /projects/{project_id}/branches
GET /projects/{project_id}/databases
POST /projects/{project_id}/databases
GET /projects/{project_id}/databases/{database_id}
PUT /projects/{project_id}/databases/{database_id}
GET /projects/{project_id}/operations/{operation_id}
POST /projects/{project_id}/query
GET /projects/{project_id}/roles
POST /projects/{project_id}/roles
GET /projects/{project_id}/roles/{role_name}
DELETE /projects/{project_id}/roles/{role_name}
POST /projects/{project_id}/roles/{role_name}/reset_password
POST /projects/{project_id}/start
POST /projects/{project_id}/stop
POST /psql_session/{psql_session_id}
```
It looks fine and we probably already have clients relying on it. So we should not change it, at least for now. But most of these endpoints (if not all) are related to storage, and it can suggest us what control-plane API should look like:
GET /tenants/{tenant_id}/operations/{operation_id}
POST /tenants/{tenant_id}/query
GET /tenants/{tenant_id}/roles
POST /tenants/{tenant_id}/roles
GET /tenants/{tenant_id}/roles/{role_name}
DELETE /tenants/{tenant_id}/roles/{role_name}
POST /tenants/{tenant_id}/roles/{role_name}/reset_password
POST /tenants/{tenant_id}/start
POST /tenants/{tenant_id}/stop
POST /psql_session/{psql_session_id}
```
One of the options here is to use gRPC instead of the HTTP, which has some useful features, but there are some strong points towards using plain HTTP:
- HTTP API is easier to use for the clients
- we already have HTTP API in pageserver/safekeeper/console
- we probably want control-plane API to be similar to the console API, available in the cloud
### Getting updates from the storage
There can be some valid cases, when we would like to know what is changed in the storage. For example, console might want to know when user has queried and started compute and when compute was scaled to zero after that, to know how much user should pay for the service. Another example is to get info about reaching the disk space limits. Yet another example is to do analytics, such as how many users had at least one active project in a month.
All of the above cases can happen without using the console, just by accessing compute through the proxy.
To solve this, we can have a log of events occurring in the storage (event logs). That is very similar to operations table we have right now, the only difference is that events are immutable and we cannot change them after saving to the database. For example, we might want to have events for the following activities:
- We finished processing some HTTP API query, such as resetting the password
- We changed some state, such as started or stopped a compute
- Operation is created
- Operation is started for the first time
- Operation is failed for the first time
- Operation is finished
Once we save these events to the database, we can create HTTP API to subscribe to these events. That API can look like this:
```
GET /events/<cursor>
{
"events": [...],
"next_cursor": 123
}
```
It should be possible to replay event logs from some point of time, to get a state of almost anything from the storage services. That means that if we maintain some state in the control-plane database and we have a reason to have the same state in the console database, it is possible by polling events from the control-plane API and changing the state in the console database according to the events.
### Next steps
After implementing control-plane HTTP API and starting control-plane as a separate service, we might want to think of exploiting benefits of the new architecture, such as reorganizing test infrastructure. Possible options are listed in the [Next steps](#next-steps-1).
## Non Goals
RFC doesn’t cover the actual cloud deployment scripts and schemas, such as terraform, ansible, k8s yaml’s and so on.
## Impacted components
Mostly console, but can also affect some storage service.
## Scalability
We should support starting several instances of the new control-plane service at the same time.
At the same time, it should be possible to use only single instance of control-plane, which can be useful for local tests.
## Security implications
New control-plane service is an internal service, so no external requests can reach it. But at the same time, it contains API to do absolutely anything with any of the tenants. That means that bad internal actor can potentially read and write all of the tenants. To make this safer, we can have one of these:
- Simple option is to protect all requests with a single private key, so that no one can make requests without having that one key.
- Another option is to have a separate token for every tenant and store these tokens in another secure place. This way it’s harder to access all tenants at once, because they have the different tokens.
## Alternative implementation
There was an idea to create a k8s operator for managing storage services and computes, but author of this RFC is not really familiar with it.
Regarding less alternative ideas, there are another options for the name of the new control-plane service:
- storage-ctl
- cloud
- cloud-ctl
## Pros/cons of proposed approaches (TODO)
Pros:
- All storage features are completely open-source
- Better tests coverage, less difference between cloud and local setups
- Easier to develop storage and cloud features, because there is no need to setup console for that
- Easier to deploy storage-only services to the any cloud
Cons:
- All storage features are completely open-source
- Distributed services mean more code to connect different services and potential network issues
- Console needs to have a dependency on storage API, there can be complications with developing new feature in a branch
- More code to JOIN data from different services (console and control-plane)
## Definition of Done
We have a new control-plane service running in the k8s. Source code for that control-plane service is located in the open-source neon repo.
## Next steps
After we’ve reached DoD, we can make further improvements.
First thing that can benefit from the split is local testing. The same control-plane service can implement starting computes as a local processes instead of k8s deployments. If it will also support starting pageservers/safekeepers/proxy for the local setup, then it can completely replace `./neon_local` binary, which is currently used for testing. The local testing environment can look like this:
```
┌─────────────────────┐ ┌───────────────────────┐
│ │ │ Storage (local) │
│ control-plane db │ │ │
│ (local process) │ │ - safekeepers │
│ │ │ - pageservers │
└──────────▲──────────┘ │ │
│ │ Dependencies │
┌──────────┴──────────┐ │ │
│ │ │ - etcd │
│ control-plane ├────►│ - S3 │
│ (local process) │ │ - more? │
│ │ │ │
└──────────┬──────────┘ └───────────────────────┘
▲ │ ▲
│ │ │
│ │ ┌───────────┴───────────┐
│ │ │ │
│ └───────────────►│ computes │
│ │ (local processes) │
│ │ │
┌──────┴──────────────┐ └───────────────────────┘
│ │ ▲
│ proxy │ │
│ (local process) ├─────────────────┘
│ │
└─────────────────────┘
```
The key thing here is that control-plane local service have the same API and almost the same implementation as the one deployed in the k8s. This allows to run the same e2e tests against both cloud and local setups.
For the python test_runner tests everything can stay mostly the same. To do that, we just need to replace `./neon_local` cli commands with API calls to the control-plane.
The benefit here will be in having fast local tests that are really close to our cloud setup. Bugs in k8s queries are still cannot be found when running computes as a local processes, but it should be really easy to start k8s locally (for example in k3s) and run the same tests with control-plane connected to the local k8s.
Talking about console and UI tests, after the split there should be a way to test these without spinning up all the storage locally. New control-plane service has a well-defined API, allowing us to mock it. This way we can create UI tests to verify the right calls are issued after specific UI interactions and verify that we render correct messages when API returns errors.
Currently we dont delete pageserver part of the data from s3 when project is deleted. (The same is true for safekeepers, but this outside of the scope of this RFC).
Currently we don't delete pageserver part of the data from s3 when project is deleted. (The same is true for safekeepers, but this outside of the scope of this RFC).
This RFC aims to spin a discussion to come to a robust deletion solution that wont put us in into a corner for features like postponed deletion (when we keep data for user to be able to restore a project if it was deleted by accident)
@@ -75,9 +75,9 @@ Remote one is needed for cases when pageserver is lost during deletion so other
Why local mark file is needed?
If we dont have one, we have two choices, delete local data before deleting the remote part or do that after.
If we don't have one, we have two choices, delete local data before deleting the remote part or do that after.
If we delete local data before remote then during restart pageserver wont pick up remote tenant at all because nothing is available locally (pageserver looks for remote conuterparts of locally available tenants).
If we delete local data before remote then during restart pageserver wont pick up remote tenant at all because nothing is available locally (pageserver looks for remote counterparts of locally available tenants).
If we delete local data after remote then at the end of the sequence when remote mark file is deleted if pageserver restart happens then the state is the same to situation when pageserver just missing data on remote without knowing the fact that this data is intended to be deleted. In this case the current behavior is upload everything local-only to remote.
@@ -145,7 +145,7 @@ sequenceDiagram
CP->>PS: Retry delete tenant
PS->>CP: Not modified
else Mark is missing
note over PS: Continue to operate the tenant as if deletion didnt happen
note over PS: Continue to operate the tenant as if deletion didn't happen
note over CP: Eventually console should <br> retry delete request
@@ -168,7 +168,7 @@ sequenceDiagram
PS->>CP: True
```
Similar sequence applies when both local and remote marks were persisted but Control Plane still didnt receive a response.
Similar sequence applies when both local and remote marks were persisted but Control Plane still didn't receive a response.
If pageserver crashes after both mark files were deleted then it will reply to control plane status poll request with 404 which should be treated by control plane as success.
@@ -187,7 +187,7 @@ If pageseserver is lost then the deleted tenant should be attached to different
##### Restrictions for tenant that is in progress of being deleted
I propose to add another state to tenant/timeline - PendingDelete. This state shouldnt allow executing any operations aside from polling the deletion status.
I propose to add another state to tenant/timeline - PendingDelete. This state shouldn't allow executing any operations aside from polling the deletion status.
#### Summary
@@ -237,7 +237,7 @@ New branch gets created
PS1 starts up (is it possible or we just recycle it?)
PS1 is unaware of the new branch. It can either fall back to s3 ls, or ask control plane.
So here comes the dependency of storage on control plane. During restart storage needs to know which timelines are valid for operation. If there is nothing on s3 that can answer that question storage neeeds to ask control plane.
So here comes the dependency of storage on control plane. During restart storage needs to know which timelines are valid for operation. If there is nothing on s3 that can answer that question storage needs to ask control plane.
### Summary
@@ -250,7 +250,7 @@ Cons:
Pros:
- Easier to reason about if you dont have to account for pageserver restarts
- Easier to reason about if you don't have to account for pageserver restarts
### Extra notes
@@ -262,7 +262,7 @@ Delayed deletion can be done with both approaches. As discussed with Anna (@step
After discussion in comments I see that we settled on two options (though a bit different from ones described in rfc). First one is the same - pageserver owns as much as possible. The second option is that pageserver owns markers thing, but actual deletion happens in control plane by repeatedly calling ls + delete.
To my mind the only benefit of the latter approach is possible code reuse between safekeepers and pageservers. Otherwise poking around integrating s3 library into control plane, configuring shared knowledge abouth paths in s3 - are the downsides. Another downside of relying on control plane is the testing process. Control plane resides in different repository so it is quite hard to test pageserver related changes there. e2e test suite there doesnt support shutting down pageservers, which are separate docker containers there instead of just processes.
To my mind the only benefit of the latter approach is possible code reuse between safekeepers and pageservers. Otherwise poking around integrating s3 library into control plane, configuring shared knowledge about paths in s3 - are the downsides. Another downside of relying on control plane is the testing process. Control plane resides in different repository so it is quite hard to test pageserver related changes there. e2e test suite there doesn't support shutting down pageservers, which are separate docker containers there instead of just processes.
With pageserver owning everything we still give the retry logic to control plane but its easier to duplicate if needed compared to sharing inner s3 workings. We will have needed tests for retry logic in neon repo.
At this point it is not possible to restore the state from index, it contains L2 which
is no longer available in s3 and doesnt contain L3 added by compaction by the
is no longer available in s3 and doesn't contain L3 added by compaction by the
first pageserver. So if any of the pageservers restart, initial sync will fail
(or in on-demand world it will fail a bit later during page request from
missing layer)
@@ -171,13 +171,13 @@ sequenceDiagram
Another problem is a possibility of concurrent branch creation calls.
I e during migration create_branch can be called on old pageserver and newly created branch wont be seen on new pageserver. Prior art includes prototyping an approach of trying to mirror such branches, but currently it lost its importance, because now attach is fast because we dont need to download all data, and additionally to the best of my knowledge of control plane internals (cc @ololobus to confirm) operations on one project are executed sequentially, so it is not possible to have such case. So branch create operation will be executed only when relocation is completed. As a safety measure we can forbid branch creation for tenants that are in readonly remote state.
I e during migration create_branch can be called on old pageserver and newly created branch wont be seen on new pageserver. Prior art includes prototyping an approach of trying to mirror such branches, but currently it lost its importance, because now attach is fast because we don't need to download all data, and additionally to the best of my knowledge of control plane internals (cc @ololobus to confirm) operations on one project are executed sequentially, so it is not possible to have such case. So branch create operation will be executed only when relocation is completed. As a safety measure we can forbid branch creation for tenants that are in readonly remote state.
## Simplistic approach
The difference of simplistic approach from one described above is that it calls ignore on source tenant first and then calls attach on target pageserver. Approach above does it in opposite order thus opening a possibility for race conditions we strive to avoid.
The approach largely follows this guide: <https://github.com/neondatabase/cloud/wiki/Cloud:-Ad-hoc-tenant-relocation>
The approach largely follows this guide: <https://www.notion.so/neondatabase/Cloud-Ad-hoc-tenant-relocation-f687474f7bfc42269e6214e3acba25c7>
// the most recently started txn's id; only most recently sarted can win
// the most recently started txn's id; only most recently started can win
next_winner_txn: Option<TxnId>,
}
struct Transaction {
@@ -186,7 +186,7 @@ A transaction T in state Committed has subsequent transactions that may or may n
So, for garbage collection, we need to assess transactions in state Committed and RejectAcknowledged:
- Commited: delete objects on the deadlist.
- Committed: delete objects on the deadlist.
- We don’t need a LIST request here, the deadlist is sufficient. So, it’s really cheap.
- This is **not true MVCC garbage collection**; by deleting the objects on Committed transaction T ’s deadlist, we might delete data referenced by other transactions that were concurrent with T, i.e., they started while T was still open. However, the fact that T is committed means that the other transactions are RejectPending or RejectAcknowledged, so, they don’t matter. Pageservers executing these doomed RejectPending transactions must handle 404 for GETs gracefully, e.g., by trying to commit txn so they observe the rejection they’re destined to get anyways. 404’s for RejectAcknowledged is handled below.
- RejectAcknowledged: delete all objects created in that txn, and discard deadlists.
@@ -242,15 +242,15 @@ If a pageserver is unresponsive from Control Plane’s / Compute’s perspective
At this point, availability is restored and user pain relieved.
What’s left is to somehow close the doomed transaction of the unresponsive pageserver, so that it beomes RejectAcknowledged, and GC can make progress. Since S3 is cheap, we can afford to wait a really long time here, especially if we put a soft bound on the amount of data a transaction may produce before it must commit. Procedure:
What’s left is to somehow close the doomed transaction of the unresponsive pageserver, so that it becomes RejectAcknowledged, and GC can make progress. Since S3 is cheap, we can afford to wait a really long time here, especially if we put a soft bound on the amount of data a transaction may produce before it must commit. Procedure:
1. Ensure the unresponsive pageserver is taken out of rotation for new attachments. That probably should happen as part of the routine above.
2. Make a human operator investigate decide what to do (next morning, NO ONCALL ALERT):
1. Inspect the instance, investigate logs, understand root cause.
2. Try to re-establish connectivity between pageserver and Control Plane so that pageserver can retry commits, get rejected, ack rejection ⇒ enable GC.
3. Use below procedure to decomission pageserver.
3. Use below procedure to decommission pageserver.
### Decomissioning A Pageserver (Dead or Alive-but-Unrespsonive)
### Decommissioning A Pageserver (Dead or Alive-but-Unresponsive)
The solution, enabled by this proposal:
@@ -310,7 +310,7 @@ Issues that we discussed:
1. In abstract terms, this proposal provides a linearized history for a given S3 prefix.
2. In concrete terms, this proposal provides a linearized history per tenant.
3. There can be multiple writers at a given time, but only one of them will win to become part of the linearized history.
4.************************************************************************************Alternative ideas mentioned during meetings that should be turned into a written prospoal like this one:************************************************************************************
4.************************************************************************************Alternative ideas mentioned during meetings that should be turned into a written proposal like this one:************************************************************************************
1.@Dmitry Rodionov : having linearized storage of index_part.json in some database that allows serializable transactions / atomic compare-and-swap PUT
2.@Dmitry Rodionov :
3.@Stas : something like this scheme, but somehow find a way to equate attachment duration with transaction duration, without losing work if pageserver dies months after attachment.
@@ -54,7 +54,7 @@ If the compaction algorithm doesn't change between the two compaction runs, is d
*However*:
1. the file size of the overwritten L1s may not be identical, and
2. the bit pattern of the overwritten L1s may not be identical, and,
3. in the future, we may want to make the compaction code non-determinstic, influenced by past access patterns, or otherwise change it, resulting in L1 overwrites with a different set of delta records than before the overwrite
3. in the future, we may want to make the compaction code non-deterministic, influenced by past access patterns, or otherwise change it, resulting in L1 overwrites with a different set of delta records than before the overwrite
The items above are a problem for the [split-brain protection RFC](https://github.com/neondatabase/neon/pull/4919) because it assumes that layer files in S3 are only ever deleted, but never replaced (overPUTted).
@@ -63,7 +63,7 @@ But node B based its world view on the version of node A's `index_part.json` fro
That earlier `index_part.json`` contained the file size of the pre-overwrite L1.
If the overwritten L1 has a different file size, node B will refuse to read data from the overwritten L1.
Effectively, the data in the L1 has become inaccessible to node B.
If node B already uploaded an index part itself, all subsequent attachments will use node B's index part, and run into the same probem.
If node B already uploaded an index part itself, all subsequent attachments will use node B's index part, and run into the same problem.
If we ever introduce checksums instead of checking just the file size, then a mismatching bit pattern (2) will cause similar problems.
@@ -121,7 +121,7 @@ Multi-object changes that previously created and removed files in timeline dir a
* atomic `index_part.json` update in S3, as per guarantee that S3 PUT is atomic
* local timeline dir state:
* irrelevant for layer map content => irrelevant for atomic updates / crash consistency
* if we crash after index part PUT, local layer files will be used, so, no on-demand downloads neede for them
* if we crash after index part PUT, local layer files will be used, so, no on-demand downloads needed for them
* if we crash before index part PUT, local layer files will be deleted
## Trade-Offs
@@ -140,7 +140,7 @@ Assuming upload queue allows for unlimited queue depth (that's what it does toda
* wal ingest: currently unbounded
* L0 => L1 compaction: CPU time proportional to `O(sum(L0 size))` and upload work proportional to `O()`
* Compaction threshold is 10 L0s and each L0 can be up to 256M in size. Target size for L1 is 128M.
* In practive, most L0s are tiny due to 10minute `DEFAULT_CHECKPOINT_TIMEOUT`.
* In practice, most L0s are tiny due to 10minute `DEFAULT_CHECKPOINT_TIMEOUT`.
* image layer generation: CPU time `O(sum(input data))` + upload work `O(sum(new image layer size))`
* I have no intuition how expensive / long-running it is in reality.
* gc: `update_gc_info`` work (not substantial, AFAIK)
@@ -158,7 +158,7 @@ Pageserver crashes are very rare ; it would likely be acceptable to re-do the lo
However, regular pageserver restart happen frequently, e.g., during weekly deploys.
In general, pageserver restart faces the problem of tenants that "take too long" to shut down.
They are a problem because other tenants that shut down quickly are unavailble while we wait for the slow tenants to shut down.
They are a problem because other tenants that shut down quickly are unavailable while we wait for the slow tenants to shut down.
We currently allot 10 seconds for graceful shutdown until we SIGKILL the pageserver process (as per `pageserver.service` unit file).
A longer budget would expose tenants that are done early to a longer downtime.
A short budget would risk throwing away more work that'd have to be re-done after restart.
To guarantee uniqueness, the unqiue number is a sequence number, stored in `index_part.json`.
To guarantee uniqueness, the unique number is a sequence number, stored in `index_part.json`.
This alternative does not solve atomic layer map updates.
In our crash-during-compaction scenario above, the compaction run after the crash will not overwrite the L1s, but write/PUT new files with new sequence numbers.
@@ -246,11 +246,11 @@ We'd need to write a deduplication pass that checks if perfectly overlapping lay
However, this alternative is appealing because it systematically prevents overwrites at a lower level than this RFC.
So, this alternative is sufficient for the needs of the split-brain safety RFC (immutable layer files locally and in S3).
But it doesn't solve the problems with crash-during-compaction outlined earlier in this RFC, and in fact, makes it much more accute.
But it doesn't solve the problems with crash-during-compaction outlined earlier in this RFC, and in fact, makes it much more acute.
The proposed design in this RFC addresses both.
So, if this alternative sounds appealing, we should implement the proposal in this RFC first, then implement this alternative on top.
That way, we avoid a phase where the crash-during-compaction problem is accute.
That way, we avoid a phase where the crash-during-compaction problem is acute.
This RFC proposes per-tenant throttling of GetPage@LSN requests inside Pageserver
and the interactions with its client, i.e., the neon_smgr component in Compute.
The result of implementing & executing this RFC will be a fleet-wide upper limit for
**"the highest GetPage/second that Pageserver can support for a single tenant/shard"**.
## Background
### GetPage@LSN Request Flow
Pageserver exposes its `page_service.rs` as a libpq listener.
The Computes' `neon_smgr` module connects to that libpq listener.
Once a connection is established, the protocol allows Compute to request page images at a given LSN.
We call these requests GetPage@LSN requests, or GetPage requests for short.
Other request types can be sent, but these are low traffic compared to GetPage requests
and are not the concern of this RFC.
Pageserver associates one libpq connection with one tokio task.
Per connection/task, the pq protocol is handled by the common `postgres_backend` crate.
Its `run_message_loop` function invokes the `page_service` specific `impl<IO> postgres_backend::Handler<IO> for PageServerHandler`.
Requests are processed in the order in which they arrive via the TCP-based pq protocol.
So, there is no concurrent request processing within one connection/task.
There is a degree of natural pipelining:
Compute can "fill the pipe" by sending more than one GetPage request into the libpq TCP stream.
And Pageserver can fill the pipe with responses in the other direction.
Both directions are subject to the limit of tx/rx buffers, nodelay, TCP flow control, etc.
### GetPage@LSN Access Pattern
The Compute has its own hierarchy of caches, specifically `shared_buffers` and the `local file cache` (LFC).
Compute only issues GetPage requests to Pageserver if it encounters a miss in these caches.
If the working set stops fitting into Compute's caches, requests to Pageserver increase sharply -- the Compute starts *thrashing*.
## Motivation
In INC-69, a tenant issued 155k GetPage/second for a period of 10 minutes and 60k GetPage/second for a period of 3h,
then dropping to ca 18k GetPage/second for a period of 9h.
We noticed this because of an internal GetPage latency SLO burn rate alert, i.e.,
the request latency profile during this period significantly exceeded what was acceptable according to the internal SLO.
Sadly, we do not have the observability data to determine the impact of this tenant on other tenants on the same tenants.
However, here are some illustrative data points for the 155k period:
The tenant was responsible for >= 99% of the GetPage traffic and, frankly, the overall activity on this Pageserver instance.
We were serving pages at 10 Gb/s (`155k x 8 kbyte (PAGE_SZ) per second is 1.12GiB/s = 9.4Gb/s.`)
The CPU utilization of the instance was 75% user+system.
Pageserver page cache served 1.75M accesses/second at a hit rate of ca 90%.
The hit rate for materialized pages was ca. 40%.
Curiously, IOPS to the Instance Store NVMe were very low, rarely exceeding 100.
The fact that the IOPS were so low / the materialized page cache hit rate was so high suggests that **this tenant's compute's caches were thrashing**.
The compute was of type `k8s-pod`; hence, auto-scaling could/would not have helped remediate the thrashing by provisioning more RAM.
The consequence was that the **thrashing translated into excessive GetPage requests against Pageserver**.
My claim is that it was **unhealthy to serve this workload at the pace we did**:
* it is likely that other tenants were/would have experienced high latencies (again, we sadly don't have per-tenant latency data to confirm this)
* more importantly, it was **unsustainable** to serve traffic at this pace for multiple reasons:
* **predictability of performance**: when the working set grows, the pageserver materialized page cache hit rate drops.
At some point, we're bound by the EC2 Instance Store NVMe drive's IOPS limit.
The result is an **uneven** performance profile from the Compute perspective.
* **economics**: Neon currently does not charge for IOPS, only capacity.
**We cannot afford to undercut the market in IOPS/$ this drastically; it leads to adverse selection and perverse incentives.**
For example, the 155k IOPS, which we served for 10min, would cost ca. 6.5k$/month when provisioned as an io2 EBS volume.
Even the 18k IOPS, which we served for 9h, would cost ca. 1.1k$/month when provisioned as an io2 EBS volume.
We charge 0$.
It could be economically advantageous to keep using a low-DRAM compute because Pageserver IOPS are fast enough and free.
Note: It is helpful to think of Pageserver as a disk, because it's precisely where `neon_smgr` sits:
vanilla Postgres gets its pages from disk, Neon Postgres gets them from Pageserver.
So, regarding the above performance & economic arguments, it is fair to say that we currently provide an "as-fast-as-possible-IOPS" disk that we charge for only by capacity.
## Solution: Throttling GetPage Requests
**The consequence of the above analysis must be that Pageserver throttles GetPage@LSN requests**.
That is, unless we want to start charging for provisioned GetPage@LSN/second.
Throttling sets the correct incentive for a thrashing Compute to scale up its DRAM to the working set size.
Neon Autoscaling will make this easy, [eventually](https://github.com/neondatabase/neon/pull/3913).
## The Design Space
What that remains is the question about *policy* and *mechanism*:
**Policy** concerns itself with the question of what limit applies to a given connection|timeline|tenant.
Candidates are:
* hard limit, same limit value per connection|timeline|tenant
* Per-tenant will provide an upper bound for the impact of a tenant on a given Pageserver instance.
This is a major operational pain point / risk right now.
* hard limit, configurable per connection|timeline|tenant
* This outsources policy to console/control plane, with obvious advantages for flexible structuring of what service we offer to customers.
* Note that this is not a mechanism to guarantee a minium provisioned rate, i.e., this is not a mechanism to guarantee a certain QoS for a tenant.
* fair share among active connections|timelines|tenants per instance
* example: each connection|timeline|tenant gets a fair fraction of the machine's GetPage/second capacity
* NB: needs definition of "active", and knowledge of available GetPage/second capacity in advance
* ...
Regarding **mechanism**, it's clear that **backpressure** is the way to go.
However, we must choose between
* **implicit** backpressure through pq/TCP and
* **explicit** rejection of requests + retries with exponential backoff
Further, there is the question of how throttling GetPage@LSN will affect the **internal GetPage latency SLO**:
where do we measure the SLI for Pageserver's internal getpage latency SLO? Before or after the throttling?
And when we eventually move the measurement point into the Computes (to avoid coordinated omission),
how do we avoid counting throttling-induced latency toward the internal getpage latency SLI/SLO?
## Scope Of This RFC
**This RFC proposes introducing a hard GetPage@LSN/second limit per tenant, with the same value applying to each tenant on a Pageserver**.
This proposal is easy to implement and significantly de-risks operating large Pageservers,
based on the assumption that extremely-high-GetPage-rate-episodes like the one from the "Motivation" section are uncorrelated between tenants.
For example, suppose we pick a limit that allows up to 10 tenants to go at limit rate.
Suppose our Pageserver can serve 100k GetPage/second total at a 100% page cache miss rate.
If each tenant gets a hard limit of 10k GetPage/second, we can serve up to 10 tenants at limit speed without latency degradation.
The mechanism for backpressure will be TCP-based implicit backpressure.
The compute team isn't concerned about prefetch queue depth.
Pageserver will implement it by delaying the reading of requests from the libpq connection(s).
The rate limit will be implemented using a per-tenant token bucket.
The bucket will be be shared among all connections to the tenant.
The bucket implementation supports starvation-preventing `await`ing.
The current candidate for the implementation is [`leaky_bucket`](https://docs.rs/leaky-bucket/).
The getpage@lsn benchmark that's being added in https://github.com/neondatabase/neon/issues/5771
can be used to evaluate the overhead of sharing the bucket among connections of a tenant.
A possible technique to mitigate the impact of sharing the bucket would be to maintain a buffer of a few tokens per connection handler.
Regarding metrics / the internal GetPage latency SLO:
we will measure the GetPage latency SLO _after_ the throttler and introduce a new metric to measure the amount of throttling, quantified by:
- histogram that records the tenants' observations of queue depth before they start waiting (one such histogram per pageserver)
- histogram that records the tenants' observations of time spent waiting (one such histogram per pageserver)
Further observability measures:
- an INFO log message at frequency 1/min if the tenant/timeline/connection was throttled in that last minute.
The message will identify the tenant/timeline/connection to allow correlation with compute logs/stats.
Rollout will happen as follows:
- deploy 1: implementation + config: disabled by default, ability to enable it per tenant through tenant_conf
- experimentation in staging and later production to study impact & interaction with auto-scaling
- determination of a sensible global default value
- the value will be chosen as high as possible ...
- ... but low enough to work towards this RFC's goal that one tenant should not be able to dominate a pageserver instance.
- deploy 2: implementation fixes if any + config: enabled by default with the aforementioned global default
- reset of the experimental per-tenant overrides
- gain experience & lower the limit over time
- we stop lowering the limit as soon as this RFC's goal is achieved, i.e.,
once we decide that in practice the chosen value sufficiently de-risks operating large pageservers
The per-tenant override will remain for emergencies and testing.
But since Console doesn't preserve it during tenant migrations, it isn't durably configurable for the tenant.
Toward the upper layers of the Neon stack, the resulting limit will be
**"the highest GetPage/second that Pageserver can support for a single tenant"**.
### Rationale
We decided against error + retry because of worries about starvation.
## Future Work
Enable per-tenant emergency override of the limit via Console.
Should be part of a more general framework to specify tenant config overrides.
**NB:** this is **not** the right mechanism to _sell_ different max GetPage/second levels to users,
or _auto-scale_ the GetPage/second levels. Such functionality will require a separate RFC that
concerns itself with GetPage/second capacity planning.
Compute-side metrics for GetPage latency.
Back-channel to inform Compute/Autoscaling/ControlPlane that the project is being throttled.
Compute-side neon_smgr improvements to avoid sending the same GetPage request multiple times if multiple backends experience a cache miss.
Dealing with read-only endpoints: users use read-only endpoints to scale reads for a single tenant.
Possibly there are also assumptions around read-only endpoints not affecting the primary read-write endpoint's performance.
With per-tenant rate limiting, we will not meet that expectation.
However, we can currently only scale per tenant.
Soon, we will have sharding (#5505), which will apply the throttling on a per-shard basis.
But, that's orthogonal to scaling reads: if many endpoints hit one shard, they share the same throttling limit.
To solve this properly, I think we'll need replicas for tenants / shard.
To performance-isolate a tenant's endpoints from each other, we'd then route them to different replicas.
Each of these `Timeline::get` calls must traverse the layer map to gather reconstruct data (`Timeline::get_reconstruct_data`) for the requested page number (`blknum` in the example).
For each layer visited by layer map traversal, we do a `DiskBtree` point lookup.
If it's negative (no entry), we resume layer map traversal.
If it's positive, we collect the result in our reconstruct data bag.
If the reconstruct data bag contents suffice to reconstruct the page, we're done with `get_reconstruct_data` and move on to walredo.
Otherwise, we resume layer map traversal.
Doing this many `Timeline::get` calls is quite inefficient because:
1. We do the layer map traversal repeatedly, even if, e.g., all the data sits in the same image layer at the bottom of the stack.
2. We may visit many DiskBtree inner pages multiple times for point lookup of different keys.
This is likely particularly bad for L0s which span the whole key space and hence must be visited by layer map traversal, but
may not contain the data we're looking for.
3. Anecdotally, keys adjacent in keyspace and written simultaneously also end up physically adjacent in the layer files [^1].
So, to provide the reconstruct data for N adjacent keys, we would actually only _need_ to issue a single large read to the filesystem, instead of the N reads we currently do.
The filesystem, in turn, ideally stores the layer file physically contiguously, so our large read will turn into one IOP toward the disk.
We should have a vectored aka batched aka scatter-gather style alternative API for `Timeline::get`. Having such an API unlocks:
* more efficient basebackup
* batched IO during compaction (useful for strides of unchanged pages)
* page_service: expose vectored get_page_at_lsn for compute (=> good for seqscan / prefetch)
* if [on-demand SLRU downloads](https://github.com/neondatabase/neon/pull/6151) land before vectored Timeline::get, on-demand SLRU downloads will still benefit from this API
# DoD
There is a new variant of `Timeline::get`, called `Timeline::get_vectored`.
It takes as arguments an `lsn: Lsn` and a `src: &[KeyVec]` where `struct KeyVec { base: Key, count: usize }`.
It is up to the implementor to figure out a suitable and efficient way to return the reconstructed page images.
It is sufficient to simply return a `Vec<Bytes>`, but, likely more efficient solutions can be found after studying all the callers of `Timeline::get`.
Functionally, the behavior of `Timeline::get_vectored` is equivalent to
* For each visited layer, call `Layer::get_value_reconstruct_data` at most once.
* This means, read each `DiskBtree` page at most once.
* Facilitate merging of the reads we issue to the OS and eventually NVMe.
Each of these items above represents a significant amount of work.
## Performance
Ideally, the **base performance** of a vectored get of a single page should be identical to the current `Timeline::get`.
A reasonable constant overhead over current `Timeline::get` is acceptable.
The performance improvement for the vectored use case is demonstrated in some way, e.g., using the `pagebench` basebackup benchmark against a tenant with a lot of SLRU segments.
# Implementation
High-level set of tasks / changes to be made:
- **Get clarity on API**:
- Define naive `Timeline::get_vectored` implementation & adopt it across pageserver.
- The tricky thing here will be the return type (e.g. `Vec<Bytes>` vs `impl Stream`).
- Start with something simple to explore the different usages of the API.
Then iterate with peers until we have something that is good enough.
- **Vectored Layer Map traversal**
- Vectored `LayerMap::search` (take 1 LSN and N `Key`s instead of just 1 LSN and 1 `Key`)
- Refactor `Timeline::get_reconstruct_data` to hold & return state for N `Key`s instead of 1
- The slightly tricky part here is what to do about `cont_lsn` [after we've found some reconstruct data for some keys](https://github.com/neondatabase/neon/blob/d066dad84b076daf3781cdf9a692098889d3974e/pageserver/src/tenant/timeline.rs#L2378-L2385)
but need more.
Likely we'll need to keep track of `cont_lsn` per key and continue next iteration at `max(cont_lsn)` of all keys that still need data.
- Current code calls it [here](https://github.com/neondatabase/neon/blob/d066dad84b076daf3781cdf9a692098889d3974e/pageserver/src/tenant/timeline.rs#L2378-L2384).
- Delta layers use `DiskBtreeReader::visit()` to collect the `(offset,len)` pairs for delta record blobs to load.
- Image layers use `DiskBtreeReader::get` to get the offset of the image blob to load. Underneath, that's just a `::visit()` call.
- What needs to happen to `DiskBtree::visit()`?
* Minimally
* take a single `KeyVec` instead of a single `Key` as argument, i.e., take a single contiguous key range to visit.
* Change the visit code to to invoke the callback for all values in the `KeyVec`'s key range
* This should be good enough for what we've seen when investigating basebackup slowness, because there, the key ranges are contiguous.
* Ideally:
* Take a `&[KeyVec]`, sort it;
* during Btree traversal, peek at the next `KeyVec` range to determine whether we need to descend or back out.
* NB: this should be a straight-forward extension of the minimal solution above, as we'll already be checking for "is there more key range in the requested `KeyVec`".
- **Facilitate merging of the reads we issue to the OS and eventually NVMe.**
- The `DiskBtree::visit` produces a set of offsets which we then read from a `VirtualFile` [here](https://github.com/neondatabase/neon/blob/292281c9dfb24152b728b1a846cc45105dac7fe0/pageserver/src/tenant/storage_layer/delta_layer.rs#L772-L804)
- What needs to happen is the **vectorization of the `blob_io` interface and then the `VirtualFile` API**.
- That is tricky because
- the `VirtualFile` API, which sits underneath `blob_io`, is being touched by ongoing [io_uring work](https://github.com/neondatabase/neon/pull/5824)
- there's the question how IO buffers will be managed; currently this area relies heavily on `PageCache`, but there's controversy around the future of `PageCache`.
- The guiding principle here should be to avoid coupling this work to the `PageCache`.
- I.e., treat `PageCache` as an extra hop in the I/O chain, rather than as an integral part of buffer management.
Let's see how we can improve by doing the first three items in above list first, then revisit.
## Rollout / Feature Flags
No feature flags are required for this epic.
At the end of this epic, `Timeline::get` forwards to `Timeline::get_vectored`, i.e., it's an all-or-nothing type of change.
It is encouraged to deliver this feature incrementally, i.e., do many small PRs over multiple weeks.
That will help isolate performance regressions across weekly releases.
# Interaction With Sharding
[Sharding](https://github.com/neondatabase/neon/pull/5432) splits up the key space, see functions `is_key_local` / `key_to_shard_number`.
Just as with `Timeline::get`, callers of `Timeline::get_vectored` are responsible for ensuring that they only ask for blocks of the given `struct Timeline`'s shard.
Given that this is already the case, there shouldn't be significant interaction/interference with sharding.
However, let's have a safety check for this constraint (error or assertion) because there are currently few affordances at the higher layers of Pageserver for sharding<=>keyspace interaction.
For example, `KeySpace` is not broken up by shard stripe, so if someone naively converted the compaction code to issue a vectored get for a keyspace range it would violate this constraint.
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
