Compare commits

...

49 Commits

Author SHA1 Message Date
Alexey Kondratov
6cfaa775b0 chore(compute): Minor compute_ctl startup refactoring 2025-02-26 21:55:25 +01:00
Arpad Müller
14347630a4 ancestor detach: delete hardlinked layers on error (#10977)
Delete layers that we have hardlinked so far when there is an error in
`remote_copy`. This prevents a retry of the ancestor detach from
stumbling over already present layer files: the hardlink would fail with
an error.

If there is a crash, we already clean up during the timeline attach: we
loop over all layer files and purge all layers that are not referenced
by the `index_part.json`.

Make sure to hold the timeline gate to prevent races with
detach&attach&read from the layer file.

These cleanups aren't completely enough however, as there is code after
`prepare` as well. To handle errors there, we add a special case for
`AlreadyExists` errors during the hardlink, where we check if the layer
is an orphan, and if yes, we delete it from local disk. That is ideally
not the case we hit, as it is less clear in that scenario where the
layer came from, but it provides good defense in depth.

Related #10729
Fixes #10970
2025-02-26 16:11:15 +00:00
Erik Grinaker
86b9703f06 pageserver: set SO_KEEPALIVE on the page service socket (#10992)
## Problem

If the client connection goes dead without an explicit close (e.g. due
to network infrastructure dropping the connection) then we currently
won't detect it for a long time, which may e.g. block GetPage flushes
and keep the task running.

Touches https://github.com/neondatabase/cloud/issues/23515.

## Summary of changes

Enable `SO_KEEPALIVE` on the page service socket, to enable periodic TCP
keepalive probes. These are configured via Linux sysctls, which will be
deployed separately. By default, the first probe is sent after 2 hours,
so this doesn't have a practical effect until we change the sysctls.
2025-02-26 14:36:05 +00:00
Arseny Sher
01581f3af5 safekeeper: drop json_ctrl (#10722)
## Problem

json_ctrl.rs is an obsolete attempt to have tests with fine control of
feeding messages into safekeeper superseded by desim framework.

## Summary of changes

Drop it.
2025-02-26 13:32:37 +00:00
Arpad Müller
f94286f0c9 Upgrade compute_tools and compute_api to edition 2024 (#10983)
Updates `compute_tools` and `compute_api` crates to edition 2024. We
like to stay on the latest edition if possible. There is no functional
changes, however some code changes had to be done to accommodate the
edition's breaking changes.

The PR has three commits:

* the first commit updates the named crates to edition 2024 and appeases
`cargo clippy` by changing code.
* the second commit performs a `cargo fmt` that does some minor changes
(not many)
* the third commit performs a cargo fmt with nightly options to reorder
imports as a one-time thing. it's completely optional, but I offer it
here for the compute team to review it.

I'd like to hear opinions about the third commit, if it's wanted and
felt worth the diff or not. I think most attention should be put onto
the first commit.

Part of #10918
2025-02-26 13:12:26 +00:00
Fedor Dikarev
c2a768086d add credentials for pulling containers for the jobs (#10987)
Ref: https://github.com/neondatabase/cloud/issues/24939

## Problem
I found that we are missing authorization for some container jobs, that
will make them use anonymous pulls. It's not an issue for now, with high
enough limits, but that could be an issue when new limits introduced in
DockerHub (10 pulls / hour)

## Summary of changes
- add credentials for the jobs that run in containers
2025-02-26 12:50:06 +00:00
Vlad Lazar
622a9def6f tests: use generated record lsn instead of hardcoded one (#10990)
... and start the initial reader with the correct lsn

Closes https://github.com/neondatabase/neon/issues/10978
2025-02-26 12:47:13 +00:00
Arpad Müller
26bda17551 storcon: use the SchedulingPolicy enum in SafekeeperPersistence (#10897)
We don't want to serialize to/from string all the time, so use
`SchedulingPolicy` in `SafekeeperPersistence` via the use of a wrapper.

Stacked atop #10891
2025-02-26 12:12:50 +00:00
Folke Behrens
0d36f52a6c proxy: Record and export user-agent header (#10955)
neondatabase/cloud#24464
2025-02-26 11:39:34 +00:00
Heikki Linnakangas
40ad42d556 Silence "sudo: unable to resolve host" messages at compute startup (#10985) 2025-02-26 10:10:05 +00:00
Heikki Linnakangas
e452f2a5a3 Remove some redundant log lines at postgres startup (#10958) 2025-02-26 10:06:42 +00:00
Heikki Linnakangas
43b109af69 compute_ctl: Add more detailed tracing spans to startup subroutines (#10979)
In local dev environment, these steps take around 100 ms, and they are
in the critical path of a compute startup on a compute pool hit. I don't
know if it's like that in production, but as first step, add tracing
spans to the functions so that they can be measured more easily.
2025-02-26 09:51:07 +00:00
Arthur Petukhovsky
3684162d9f Bump vm-builder v0.37.1 -> v0.42.2 (#10981)
Bump version to pick up changes introduced in
https://github.com/neondatabase/autoscaling/pull/1286

It's better to have a compute release for this change first, because:
- vm-runner changes kernel loglevel from 7 to 6
- vm-builder has a change to bring it back to 7 after startup

Previous update: https://github.com/neondatabase/neon/pull/10015
2025-02-26 09:19:19 +00:00
Arpad Müller
920040e402 Update storage components to edition 2024 (#10919)
Updates storage components to edition 2024. We like to stay on the
latest edition if possible. There is no functional changes, however some
code changes had to be done to accommodate the edition's breaking
changes.

The PR has two commits:

* the first commit updates storage crates to edition 2024 and appeases
`cargo clippy` by changing code. i have accidentially ran the formatter
on some files that had other edits.
* the second commit performs a `cargo fmt`

I would recommend a closer review of the first commit and a less close
review of the second one (as it just runs `cargo fmt`).

part of https://github.com/neondatabase/neon/issues/10918
2025-02-25 23:51:37 +00:00
Konstantin Knizhnik
dc975d554a Incremenet getpage histogram in prefetch_lookup (#10965)
## Problem

PR https://github.com/neondatabase/neon/pull/10442 added prefetch_lookup
function.
It changed handling of getpage requests in compute.

Before:
1. Lookup in LFC (return if found)
2. Register prefetch buffer
3. Wait prefetch result (increment getpage_hist)
Now:

1. Lookup prefetch ring (return if prefetch request is already
completed)
2. Lookup in LFC (return if found)
3. Register prefetch buffer
4. Wait prefetch result (increment getpage_hist)

So if prefetch result is already available, then get page histogram is
not incremented.
It case failure of some test_throughtput benchmarks:
https://neondb.slack.com/archives/C033RQ5SPDH/p1740425527249499

## Summary of changes

Increment getpage histogram in `prefetch_lookup`

Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
2025-02-25 19:51:38 +00:00
Suhas Thalanki
d05606252d fix: only showing LSN for static computes in neon endpoint list (#10931)
## Problem

`neon endpoint list` shows a different LSN than what the state of the
replica is. This is mainly down to what we define as LSN in this output.
If we define it as the LSN that a compute was started with, it only
makes sense to show it for static computes.

## Summary of changes

Removed the output of `last_record_lsn` for primary/hot standby
computes.

Closes: https://github.com/neondatabase/neon/issues/5825

---------

Co-authored-by: Tristan Partin <tristan@neon.tech>
2025-02-25 19:26:14 +00:00
Alex Chi Z.
c69ebb4486 fix(ci): extend timeout to 75min (#10963)
60min is not enough for debug builds

Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-02-25 17:37:23 +00:00
a-masterov
1fb2faab5b Rename the patch files for the semver test (#10966)
## Problem
The patch for `semver` extensions relies on `PG_VERSION` environment
variable. The files were named without the letter `v` so script cannot
find them.
## Summary of changes
The patch files were renamed.
2025-02-25 16:00:43 +00:00
Alex Chi Z.
015092d259 feat(pageserver): add automatic trigger for gc-compaction (#10798)
## Problem

part of https://github.com/neondatabase/neon/issues/9114

## Summary of changes

Add the auto trigger for gc-compaction. It computes two values: L1 size
and L2 size. When L1 size >= initial trigger threshold, we will trigger
an initial gc-compaction. When l1_size / l2_size >=
gc_compaction_ratio_percent, we will trigger the "tiered" gc-compaction.

---------

Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-02-25 14:50:39 +00:00
Alex Chi Z.
b7fcf2c7a7 test(pageserver): add reldir v2 into tests (#10750)
## Problem

We have `test_perf_many_relations` but it only runs on remote clusters,
and we cannot directly modify tenant config. Therefore, I patched one of
the current tests to benchmark relv2 performance.

close https://github.com/neondatabase/neon/issues/9986

## Summary of changes

* Add `v1/v2` selector to `test_tx_abort_with_many_relations`.

---------

Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-02-25 14:50:22 +00:00
Erik Grinaker
8deeddd4f0 pageserver: ignore CollectKeySpaceError::Cancelled during compaction (#10968)
This pops up a few times during deployment. Not sure why it fires
without `self.cancel` being cancelled, but could be e.g. ancestor
timelines or sth.
2025-02-25 14:49:41 +00:00
a-masterov
f78ac44748 Use the Dockerfile COPY instead of docker cp (#10943)
## Problem
We use `docker cp` to copy the files required for the extension tests
now.
It causes problems if we run older images with the newer source tree.
## Summary of changes
Copying the files was moved to the compute Dockerfile.
2025-02-25 12:44:06 +00:00
Folke Behrens
f4fefd9f2f pre-commit: Switch to cargo fmt to handle per-crate editions (#10969)
cargo knows what edition each crate uses.
2025-02-25 12:29:27 +00:00
Konstantin Knizhnik
8f82c661d4 Move neon_pgstat_file_size_limit to the extension (#10959)
## Problem

PG14 uses separate backend for stats collector having no access to
shaerd memory.
As far as AUX mechanism requires access to shared memory, persisting
pgstat.stat file
is not supported at pg14. And so there is no definition of
`neon_pgstat_file_size_limit`
variable. It makes it impossible to provide same config for all Postgres
version.

## Summary of changes

Move neon_pgstat_file_size_limit to Neon extension.

Postgres submodules PR:
https://github.com/neondatabase/postgres/pull/587
https://github.com/neondatabase/postgres/pull/588
https://github.com/neondatabase/postgres/pull/589

---------

Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
Co-authored-by: Tristan Partin <tristan@neon.tech>
2025-02-25 12:23:04 +00:00
Arseny Sher
758f597280 compute <-> sk protocol v3 (#10647)
## Problem

As part of https://github.com/neondatabase/neon/issues/8614 we need to
pass membership configurations between compute and safekeepers.

## Summary of changes

Add version 3 of the protocol carrying membership configurations.
Greeting message in both sides gets full conf, and other messages
generation number only. Use protocol bump to include other accumulated
changes:
- stop packing whole structs on the wire as is;
- make the tag u8 instead of u64;
- send all ints in network order;
- drop proposer_uuid, we can pass it in START_WAL_PUSH and it wasn't
much useful anyway.
Per message changes, apart from mconf:
- ProposerGreeting: tenant / timeline id is sent now as hex cstring.
Remove proto version, it is passed outside in START_WAL_PUSH. Remove
postgres timeline, it is unused. Reorder fields a bit.
- AcceptorGreeting: reorder fields
- VoteResponse: timeline_start_lsn is removed. It can be taken from
first member of term history, and later we won't need it at all when all
timelines will be explicitly created. Vote itself is u8 instead of u64.
- ProposerElected: timeline_start_lsn is removed for the same reasons.
- AppendRequest: epoch_start_lsn removed, it is known from term history
in ProposerElected.

Both compute and sk are able to talk v2 and v3 to make rollbacks (in
case we need them) easier; neon.safekeeper_proto_version GUC sets the
client version. v2 code can be dropped later.

So far empty conf is passed everywhere, future PRs will handle them.

To test, add param to some tests choosing proto version; we want to test
both 2 and 3 until we fully migrate.

ref https://github.com/neondatabase/neon/issues/10326

---------

Co-authored-by: Arthur Petukhovsky <petuhovskiy@yandex.ru>
2025-02-25 11:56:05 +00:00
Vlad Lazar
0d9a45a475 safekeeper: invalidate start of interpreted batch on reader resets (#10951)
## Problem

The interpreted WAL reader tracks the start of the current logical
batch.
This needs to be invalidated when the reader is reset.

This bug caused a couple of WAL gap alerts in staging.

## Summary of changes

* Refactor to make it possible to write a reproducer
* Add repro unit test
* Fix by resetting the start with the reader

Related https://github.com/neondatabase/cloud/issues/23935
2025-02-25 10:21:35 +00:00
Vlad Lazar
5d17640944 storcon: send heartbeats concurrently (#10954)
## Problem

While looking at logs I noticed that heartbeats are sent sequentially.
The loop polling the UnorderedSet is at the wrong level of identation.
Instead of doing it after we have the full set, we did after each entry.

## Summary of Changes

Poll the UnorderedSet properly.
2025-02-25 09:33:08 +00:00
Erik Grinaker
6621be6b7b pageserver: tweak slow GetPage logging (#10956)
## Problem

We recently added slow GetPage request logging. However, this
unintentionally included the flush time when logging (which we already
have separate logging for). It also logs at WARN level, which is a bit
aggressive since we see this fire quite frequently.

Follows https://github.com/neondatabase/neon/pull/10906.

## Summary of changes

* Only log the request execution time, not the flush time.
* Extract a `pagestream_dispatch_batched_message()` helper.
* Rename `warn_slow()` to `log_slow()` and downgrade to INFO.
2025-02-24 22:01:14 +00:00
Heikki Linnakangas
565a9e62a1 compute: Disconnect if no response to a pageserver request is received (#10882)
We've seen some cases in production where a compute doesn't get a
response to a pageserver request for several minutes, or even more. We
haven't found the root cause for that yet, but whatever the reason is,
it seems overly optimistic to think that if the pageserver hasn't
responded for 2 minutes, we'd get a response if we just wait patiently a
little longer. More likely, the pageserver is dead or there's some kind
of a network glitch so that the TCP connection is dead, or at least
stuck for a long time. Either way, it's better to disconnect and
reconnect. I set the default timeout to 2 minutes, which should be
enough for any GetPage request under normal circumstances, even if the
pageserver has to download several layer files from remote storage.

Make the disconnect timeout configurable. Also make the "log interval",
after which we print a message to the log configurable, so that if you
change the disconnect timeout, you can set the log timeout
correspondingly. The default log interval is still 10 s. The new GUCs
are called "neon.pageserver_response_log_timeout" and
"neon.pageserver_response_disconnect_timeout".

Includes a basic test for the log and disconnect timeouts.

Implements issue #10857
2025-02-24 20:16:37 +00:00
Peter Bendel
8fd0f89b94 rename libduckdb.so in pg_duckdb context to avoid conflict with pg_mooncake (#10915)
## Problem

Introducing pg_duckdb caused a conflict with pg_mooncake.
Both use libduckdb.so in different versions.

## Summary of changes

- Rename the libduckdb.so to libduckdb_pg_duckdb.so in the context of
pg_duckdb so that it doesn't conflict with libduckdb.so referenced by
pg_mooncake.
- use a version map to rename the duckdb symbols to a version specific
name
  - DUCKDB_1.1.3 for pg_mooncake
  - DUCKDB_1.2.0 for pg_duckdb 

For the concept of version maps see
- https://www.man7.org/conf/lca2006/shared_libraries/slide19a.html
-
https://peeterjoot.com/2019/09/20/an-example-of-linux-glibc-symbol-versioning/
- https://akkadia.org/drepper/dsohowto.pdf
2025-02-24 17:50:49 +00:00
JC Grünhage
1f0dea9a1a feat(ci): push container images to ghcr.io as well (#10945)
## Problem
There's new rate-limits coming on docker hub. To reduce our reliance on
docker hub and the problems the limits are going to cause for us, we
want to prepare for this by also pushing our container images to ghcr.io

## Summary of changes
Push our images to ghcr.io as well and not just docker hub.
2025-02-24 17:45:23 +00:00
Heikki Linnakangas
40acb0c06d Fix usage of WaitEventSetWait() with timeout (#10947)
WaitEventSetWait() returns the number of "events" that happened, and
only that many events in the WaitEvent array are updated. When the
timeout is reached, it returns 0 and does not modify the WaitEvent array
at all. We were reading 'event.events' without checking the return
value, which would be uninitialized when the timeout was hit.

No test included, as this is harmless at the moment. But this caused the
test I'm including in PR #10882 to fail. That PR changes the logic to
loop back to retry the PQgetCopyData() call if WL_SOCKET_READABLE was
set. Currently, making an extra call to PQconsumeInput() is harmless,
but with that change in logic, it turns into a busy-wait.
2025-02-24 17:15:07 +00:00
Arpad Müller
df362de0dd Reject basebackup requests for archived timelines (#10828)
For archived timelines, we would like to prevent all non-pageserver
issued getpage requests, as users are not supposed to send these.
Instead, they should unarchive a timeline before issuing any external
read traffic.

As this is non-trivial to do, at least prevent launches of new computes,
by errorring on basebackup requests for archived timelines. In #10688,
we started issuing a warning instead of an error, because an error would
mean a stuck project. Now after we can confirm the the warning is not
present in the logs for about a week, we can issue errors.

Follow-up of #10688 
Related: #9548
2025-02-24 16:38:13 +00:00
Alex Chi Z.
5fad4a4cee feat(storcon): chaos injection of force exit (#10934)
## Problem

close https://github.com/neondatabase/cloud/issues/24485

## Summary of changes

This patch adds a new chaos injection mode for the storcon. The chaos
injector reads the crontab and exits immediately at the configured time.

---------

Signed-off-by: Alex Chi Z <chi@neon.tech>
2025-02-24 15:30:21 +00:00
Arpad Müller
fdde58120c Upgrade proxy crates to edition 2024 (#10942)
This upgrades the `proxy/` crate as well as the forked libraries in
`libs/proxy/` to edition 2024.

Also reformats the imports of those forked libraries via:

```
cargo +nightly fmt -p proxy -p postgres-protocol2 -p postgres-types2 -p tokio-postgres2 -- -l --config imports_granularity=Module,group_imports=StdExternalCrate,reorder_imports=true
```

It can be read commit-by-commit: the first commit has no formatting
changes, only changes to accomodate the new edition.

Part of #10918
2025-02-24 15:26:28 +00:00
Vlad Lazar
459446fcb8 pagesever: include visible layers in heatmaps after unarchival (#10880)
## Problem

https://github.com/neondatabase/neon/pull/10788 introduced an API for
warming up attached locations
by downloading all layers in the heatmap. We intend to use it for
warming up timelines after unarchival too,
but it doesn't work. Any heatmap generated after the unarchival will not
include our timeline, so we've lost
all those layers.

## Summary of changes

Generate a cheeky heatmap on unarchival. It includes all the visible
layers.
Use that as the `PreviousHeatmap` which inputs into actual heatmap
generation.

Closes: https://github.com/neondatabase/neon/issues/10541
2025-02-24 15:21:17 +00:00
Alexander Bayandin
17724a19e6 CI(allure-reports): update dependencies and cleanup code (#10794)
## Problem

There are a bunch of minor improvements that are too small and
insignificant as is, so collecting them in one PR.

## Summary of changes
- Add runner arch to artifact name to make it easier to distinguish
files on S3
([ref](https://neondb.slack.com/archives/C059ZC138NR/p1739365938371149))
- Use `github.event.pull_request.number` instead of parsing
`$GITHUB_EVENT_PATH` file
- Update Allure CLI and `allure-pytest`
2025-02-24 15:07:14 +00:00
John Spray
2a5d7e5a78 tests: improve compat test coverage of controller-pageserver interaction (#10848)
## Problem

We failed to detect https://github.com/neondatabase/neon/pull/10845
before merging, because the tests we run with a matrix of component
versions didn't include the ones that did live migrations.

## Summary of changes

- Do a live migration during the storage controller smoke test, since
this is a pretty core piece of functionality
- Apply a compat version matrix to the graceful cluster restart test,
since this is the functionality that we most urgently need to work
across versions to make deploys work.

I expect the first CI run of this to fail, because
https://github.com/neondatabase/neon/pull/10845 isn't merged yet.
2025-02-24 12:22:22 +00:00
Conrad Ludgate
fb77f28326 feat(proxy): add direction and private link id to billing export (#10925)
ref: https://github.com/neondatabase/cloud/issues/23385

Adds a direction flag as well as private-link ID to the traffic
reporting pipeline. We do not yet actually count ingress, but we include
the flag anyway.

I have additionally moved vpce_id string parsing earlier, since we
expect it to be utf8 (ascii).
2025-02-24 11:49:11 +00:00
Heikki Linnakangas
a6f315c9c9 Remove unnecessary dependencies to synchronous 'postgres' crate (#10938)
The synchronous 'postgres' crate is just a wrapper around the async
'tokio_postgres' crate. Some places were unnecessarily using the
re-exported NoTls and Error from the synchronous 'postgres' crate, even
though they were otherwise using the 'tokio_postgres' crate. Tidy up by
using the tokio_postgres types directly.
2025-02-24 09:40:25 +00:00
Alexey Kondratov
df264380b9 fix(compute_ctl): Skip invalid DBs in PerDatabasePhase (#10910)
## Problem

After refactoring the configuration code to phases, it became a bit
fuzzy who filters out DBs that are not present in Postgres, are invalid,
or have `datallowconn = false`. The first 2 are important for the DB
dropping case, as we could be in operation retry, so DB could be already
absent in Postgres or invalid (interrupted `DROP DATABASE`).

Recent case:
https://neondb.slack.com/archives/C03H1K0PGKH/p1740053359712419

## Summary of changes

Add a common code that filters out inaccessible DBs inside
`ApplySpecPhase::RunInEachDatabase`.
2025-02-21 21:50:50 +00:00
Arpad Müller
4bbe75de8c Update vm_monitor to edition 2024 (#10916)
Updates `vm_monitor` to edition 2024. We like to stay on the latest
edition if possible. There is no functional changes, it's only changes
due to the rustfmt edition.

part of https://github.com/neondatabase/neon/issues/10918
2025-02-21 20:29:05 +00:00
Tristan Partin
c0c3ed94a9 Fix flaky test_compute_installed_extensions_metric test (#10933)
There was a race condition with compute_ctl and the metric being
collected related to whether the neon extension had been updated or not.
compute_ctl will run `ALTER EXTENSION neon UPDATE` on compute start in
the postgres database.

Fixes: https://github.com/neondatabase/neon/issues/10932

Signed-off-by: Tristan Partin <tristan@neon.tech>
2025-02-21 18:29:48 +00:00
Konstantin Knizhnik
b1d8771d5f Store prefetch results in LFC cache once as soon as they are received (#10442)
## Problem

Prefetch is performed locally, so different backers can request the same
pages form PS.
Such duplicated request increase load of page server and network
traffic.
Making prefetch global seems to be very difficult and undesirable,
because different queries can access chunks on different speed. Storing
prefetch chunks in LFC will not completely eliminate duplicates, but can
minimise such requests.

The problem with storing prefetch result in LFC is that in this case
page is not protected by share buffer lock.
So we will have to perform extra synchronisation at LFC side.

See:

https://neondb.slack.com/archives/C0875PUD0LC/p1736772890602029?thread_ts=1736762541.116949&cid=C0875PUD0LC

@MMeent implementation of prewarm:
See https://github.com/neondatabase/neon/pull/10312/


## Summary of changes

Use conditional variables to sycnhronize access to LFC entry.

---------

Co-authored-by: Konstantin Knizhnik <knizhnik@neon.tech>
2025-02-21 16:56:16 +00:00
Vlad Lazar
3e82addd64 storcon: use Duration for duration's in the storage controller tenant config (#10928)
## Problem

The storage controller treats durations in the tenant config as strings.
These are loaded from the db.
The pageserver maps these durations to a seconds only format and we
always get a mismatch compared
to what's in the db.

## Summary of changes

Treat durations as durations inside the storage controller and not as
strings.
Nothing changes in the cross service API's themselves or the way things
are stored in the db.

I also added some logging which I would have made the investigation a
10min job:
1. Reason for why the reconciliation was spawned
2. Location config diff between the observed and wanted states
2025-02-21 15:45:00 +00:00
John Spray
5e3c234edc storcon: do more concurrent optimisations (#10929)
## Problem

Now that we rely on the optimisation logic to handle fixing things up
after some tenants are in the wrong AZ (e.g. after node failure), it's
no longer appropriate to treat optimisations as an ultra-low-priority
task. We used to reflect that low priority with a very low limit on
concurrent execution, such that we would only migrate 2 things every 20
seconds.

## Summary of changes

- Increase MAX_OPTIMIZATIONS_EXEC_PER_PASS from 2 to 16
- Increase MAX_OPTIMIZATIONS_PLAN_PER_PASS from 8 to 64.

Since we recently gave user-initiated actions their own semaphore, this
should not risk starving out API requests.
2025-02-21 14:58:49 +00:00
Arpad Müller
ff3819efc7 storcon: infrastructure for safekeeper specific JWT tokens (#10905)
Safekeepers only respond to requests with the per-token scope, or the
`safekeeperdata` JWT scope. Therefore, add infrastructure in the storage
controller for safekeeper JWTs. Also, rename the ambiguous `jwt_token`
to `pageserver_jwt_token`.

Part of #9011
Related: https://github.com/neondatabase/cloud/issues/24727
2025-02-21 11:02:02 +00:00
Arpad Müller
f927ae6e15 Return a json response in scheduling_policy handler (#10904)
Return an empty json response in the `scheduling_policy` handler.

This prevents errors of the form:

```
Error: receive body: error decoding response body: EOF while parsing a value at line 1 column 0
```

when setting the scheduling policy via the `storcon_cli`.

part of #9011.
2025-02-21 11:01:57 +00:00
Heikki Linnakangas
61d385caea Split plv8 build into two parts (#10920)
Plv8 consists of two parts:
1. the V8 engine, which is built from vendored sources, and
2. the PostgreSQL extension.

Split those into two separate steps in the Dockerfile. The first step
doesn't need any PostgreSQL sources or any other files from the neon
repository, just the build tools and the upstream plv8 sources. Use the
build-deps image as the base for that step, so that the layer can be
cached and doesn't need to be rebuilt every time. This is worthwhile
because the V8 build takes a very long time.
2025-02-21 09:03:54 +00:00
429 changed files with 8976 additions and 6113 deletions

View File

@@ -14,6 +14,7 @@
!compute/
!compute_tools/
!control_plane/
!docker-compose/ext-src
!libs/
!pageserver/
!pgxn/

View File

@@ -38,9 +38,11 @@ runs:
#
- name: Set variables
shell: bash -euxo pipefail {0}
env:
PR_NUMBER: ${{ github.event.pull_request.number }}
BUCKET: neon-github-public-dev
run: |
PR_NUMBER=$(jq --raw-output .pull_request.number "$GITHUB_EVENT_PATH" || true)
if [ "${PR_NUMBER}" != "null" ]; then
if [ -n "${PR_NUMBER}" ]; then
BRANCH_OR_PR=pr-${PR_NUMBER}
elif [ "${GITHUB_REF_NAME}" = "main" ] || [ "${GITHUB_REF_NAME}" = "release" ] || \
[ "${GITHUB_REF_NAME}" = "release-proxy" ] || [ "${GITHUB_REF_NAME}" = "release-compute" ]; then
@@ -59,8 +61,6 @@ runs:
echo "LOCK_FILE=${LOCK_FILE}" >> $GITHUB_ENV
echo "WORKDIR=${WORKDIR}" >> $GITHUB_ENV
echo "BUCKET=${BUCKET}" >> $GITHUB_ENV
env:
BUCKET: neon-github-public-dev
# TODO: We can replace with a special docker image with Java and Allure pre-installed
- uses: actions/setup-java@v4
@@ -80,8 +80,8 @@ runs:
rm -f ${ALLURE_ZIP}
fi
env:
ALLURE_VERSION: 2.27.0
ALLURE_ZIP_SHA256: b071858fb2fa542c65d8f152c5c40d26267b2dfb74df1f1608a589ecca38e777
ALLURE_VERSION: 2.32.2
ALLURE_ZIP_SHA256: 3f28885e2118f6317c92f667eaddcc6491400af1fb9773c1f3797a5fa5174953
- uses: aws-actions/configure-aws-credentials@v4
if: ${{ !cancelled() }}

View File

@@ -18,9 +18,11 @@ runs:
steps:
- name: Set variables
shell: bash -euxo pipefail {0}
env:
PR_NUMBER: ${{ github.event.pull_request.number }}
REPORT_DIR: ${{ inputs.report-dir }}
run: |
PR_NUMBER=$(jq --raw-output .pull_request.number "$GITHUB_EVENT_PATH" || true)
if [ "${PR_NUMBER}" != "null" ]; then
if [ -n "${PR_NUMBER}" ]; then
BRANCH_OR_PR=pr-${PR_NUMBER}
elif [ "${GITHUB_REF_NAME}" = "main" ] || [ "${GITHUB_REF_NAME}" = "release" ] || \
[ "${GITHUB_REF_NAME}" = "release-proxy" ] || [ "${GITHUB_REF_NAME}" = "release-compute" ]; then
@@ -32,8 +34,6 @@ runs:
echo "BRANCH_OR_PR=${BRANCH_OR_PR}" >> $GITHUB_ENV
echo "REPORT_DIR=${REPORT_DIR}" >> $GITHUB_ENV
env:
REPORT_DIR: ${{ inputs.report-dir }}
- uses: aws-actions/configure-aws-credentials@v4
if: ${{ !cancelled() }}

View File

@@ -236,5 +236,5 @@ runs:
uses: ./.github/actions/allure-report-store
with:
report-dir: /tmp/test_output/allure/results
unique-key: ${{ inputs.build_type }}-${{ inputs.pg_version }}
unique-key: ${{ inputs.build_type }}-${{ inputs.pg_version }}-${{ runner.arch }}
aws-oicd-role-arn: ${{ inputs.aws-oicd-role-arn }}

View File

@@ -337,7 +337,7 @@ jobs:
- name: Pytest regression tests
continue-on-error: ${{ matrix.lfc_state == 'with-lfc' && inputs.build-type == 'debug' }}
uses: ./.github/actions/run-python-test-set
timeout-minutes: ${{ inputs.sanitizers != 'enabled' && 60 || 180 }}
timeout-minutes: ${{ inputs.sanitizers != 'enabled' && 75 || 180 }}
with:
build_type: ${{ inputs.build-type }}
test_selection: regress

View File

@@ -11,8 +11,12 @@ on:
description: AWS region to log in to. Required when pushing to ECR.
required: false
type: string
aws-account-ids:
description: Comma separated AWS account IDs to log in to for pushing to ECR. Required when pushing to ECR.
aws-account-id:
description: AWS account ID to log in to for pushing to ECR. Required when pushing to ECR.
required: false
type: string
aws-role-to-assume:
description: AWS role to assume to for pushing to ECR. Required when pushing to ECR.
required: false
type: string
azure-client-id:
@@ -31,16 +35,6 @@ on:
description: ACR registry name. Required when pushing to ACR.
required: false
type: string
secrets:
docker-hub-username:
description: Docker Hub username. Required when pushing to Docker Hub.
required: false
docker-hub-password:
description: Docker Hub password. Required when pushing to Docker Hub.
required: false
aws-role-to-assume:
description: AWS role to assume. Required when pushing to ECR.
required: false
permissions: {}
@@ -53,6 +47,7 @@ jobs:
runs-on: ubuntu-22.04
permissions:
id-token: write # Required for aws/azure login
packages: write # required for pushing to GHCR
steps:
- uses: actions/checkout@v4
with:
@@ -67,14 +62,14 @@ jobs:
uses: aws-actions/configure-aws-credentials@v4
with:
aws-region: "${{ inputs.aws-region }}"
role-to-assume: "${{ secrets.aws-role-to-assume }}"
role-to-assume: "arn:aws:iam::${{ inputs.aws-account-id }}:role/${{ inputs.aws-role-to-assume }}"
role-duration-seconds: 3600
- name: Login to ECR
if: contains(inputs.image-map, 'amazonaws.com/')
uses: aws-actions/amazon-ecr-login@v2
with:
registries: "${{ inputs.aws-account-ids }}"
registries: "${{ inputs.aws-account-id }}"
- name: Configure Azure credentials
if: contains(inputs.image-map, 'azurecr.io/')
@@ -89,11 +84,19 @@ jobs:
run: |
az acr login --name=${{ inputs.acr-registry-name }}
- name: Login to GHCR
if: contains(inputs.image-map, 'ghcr.io/')
uses: docker/login-action@v3
with:
registry: ghcr.io
username: ${{ github.repository_owner }}
password: ${{ secrets.GITHUB_TOKEN }}
- name: Log in to Docker Hub
uses: docker/login-action@v3
with:
username: ${{ secrets.docker-hub-username }}
password: ${{ secrets.docker-hub-password }}
username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
- name: Copy docker images to target registries
run: python scripts/push_with_image_map.py

View File

@@ -398,6 +398,9 @@ jobs:
runs-on: ${{ matrix.runner }}
container:
image: ${{ matrix.image }}
credentials:
username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
options: --init
# Increase timeout to 8h, default timeout is 6h

View File

@@ -722,14 +722,14 @@ jobs:
- pg: v17
debian: bookworm
env:
VM_BUILDER_VERSION: v0.37.1
VM_BUILDER_VERSION: v0.42.2
steps:
- uses: actions/checkout@v4
- name: Downloading vm-builder
run: |
curl -fL https://github.com/neondatabase/autoscaling/releases/download/$VM_BUILDER_VERSION/vm-builder -o vm-builder
curl -fL https://github.com/neondatabase/autoscaling/releases/download/$VM_BUILDER_VERSION/vm-builder-amd64 -o vm-builder
chmod +x vm-builder
- uses: neondatabase/dev-actions/set-docker-config-dir@6094485bf440001c94a94a3f9e221e81ff6b6193
@@ -750,7 +750,8 @@ jobs:
-size=2G \
-spec=compute/vm-image-spec-${{ matrix.version.debian }}.yaml \
-src=neondatabase/compute-node-${{ matrix.version.pg }}:${{ needs.tag.outputs.build-tag }} \
-dst=neondatabase/vm-compute-node-${{ matrix.version.pg }}:${{ needs.tag.outputs.build-tag }}
-dst=neondatabase/vm-compute-node-${{ matrix.version.pg }}:${{ needs.tag.outputs.build-tag }} \
-target-arch=linux/amd64
- name: Pushing vm-compute-node image
run: |
@@ -866,68 +867,72 @@ jobs:
push-neon-image-dev:
needs: [ generate-image-maps, neon-image ]
uses: ./.github/workflows/_push-to-container-registry.yml
permissions:
id-token: write # Required for aws/azure login
packages: write # required for pushing to GHCR
with:
image-map: '${{ needs.generate-image-maps.outputs.neon-dev }}'
aws-region: ${{ vars.AWS_ECR_REGION }}
aws-account-ids: "${{ vars.NEON_DEV_AWS_ACCOUNT_ID }}"
aws-account-id: "${{ vars.NEON_DEV_AWS_ACCOUNT_ID }}"
aws-role-to-assume: "gha-oidc-neon-admin"
azure-client-id: ${{ vars.AZURE_DEV_CLIENT_ID }}
azure-subscription-id: ${{ vars.AZURE_DEV_SUBSCRIPTION_ID }}
azure-tenant-id: ${{ vars.AZURE_TENANT_ID }}
acr-registry-name: ${{ vars.AZURE_DEV_REGISTRY_NAME }}
secrets:
aws-role-to-assume: "${{ vars.DEV_AWS_OIDC_ROLE_ARN }}"
docker-hub-username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
docker-hub-password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
secrets: inherit
push-compute-image-dev:
needs: [ generate-image-maps, vm-compute-node-image ]
uses: ./.github/workflows/_push-to-container-registry.yml
permissions:
id-token: write # Required for aws/azure login
packages: write # required for pushing to GHCR
with:
image-map: '${{ needs.generate-image-maps.outputs.compute-dev }}'
aws-region: ${{ vars.AWS_ECR_REGION }}
aws-account-ids: "${{ vars.NEON_DEV_AWS_ACCOUNT_ID }}"
aws-account-id: "${{ vars.NEON_DEV_AWS_ACCOUNT_ID }}"
aws-role-to-assume: "gha-oidc-neon-admin"
azure-client-id: ${{ vars.AZURE_DEV_CLIENT_ID }}
azure-subscription-id: ${{ vars.AZURE_DEV_SUBSCRIPTION_ID }}
azure-tenant-id: ${{ vars.AZURE_TENANT_ID }}
acr-registry-name: ${{ vars.AZURE_DEV_REGISTRY_NAME }}
secrets:
aws-role-to-assume: "${{ vars.DEV_AWS_OIDC_ROLE_ARN }}"
docker-hub-username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
docker-hub-password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
secrets: inherit
push-neon-image-prod:
if: github.ref_name == 'release' || github.ref_name == 'release-proxy' || github.ref_name == 'release-compute'
needs: [ generate-image-maps, neon-image, test-images ]
uses: ./.github/workflows/_push-to-container-registry.yml
permissions:
id-token: write # Required for aws/azure login
packages: write # required for pushing to GHCR
with:
image-map: '${{ needs.generate-image-maps.outputs.neon-prod }}'
aws-region: ${{ vars.AWS_ECR_REGION }}
aws-account-ids: "${{ vars.NEON_PROD_AWS_ACCOUNT_ID }}"
aws-account-id: "${{ vars.NEON_PROD_AWS_ACCOUNT_ID }}"
aws-role-to-assume: "gha-oidc-neon-admin"
azure-client-id: ${{ vars.AZURE_PROD_CLIENT_ID }}
azure-subscription-id: ${{ vars.AZURE_PROD_SUBSCRIPTION_ID }}
azure-tenant-id: ${{ vars.AZURE_TENANT_ID }}
acr-registry-name: ${{ vars.AZURE_PROD_REGISTRY_NAME }}
secrets:
aws-role-to-assume: "${{ secrets.PROD_GHA_OIDC_ROLE }}"
docker-hub-username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
docker-hub-password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
secrets: inherit
push-compute-image-prod:
if: github.ref_name == 'release' || github.ref_name == 'release-proxy' || github.ref_name == 'release-compute'
needs: [ generate-image-maps, vm-compute-node-image, test-images ]
uses: ./.github/workflows/_push-to-container-registry.yml
permissions:
id-token: write # Required for aws/azure login
packages: write # required for pushing to GHCR
with:
image-map: '${{ needs.generate-image-maps.outputs.compute-prod }}'
aws-region: ${{ vars.AWS_ECR_REGION }}
aws-account-ids: "${{ vars.NEON_PROD_AWS_ACCOUNT_ID }}"
aws-account-id: "${{ vars.NEON_PROD_AWS_ACCOUNT_ID }}"
aws-role-to-assume: "gha-oidc-neon-admin"
azure-client-id: ${{ vars.AZURE_PROD_CLIENT_ID }}
azure-subscription-id: ${{ vars.AZURE_PROD_SUBSCRIPTION_ID }}
azure-tenant-id: ${{ vars.AZURE_TENANT_ID }}
acr-registry-name: ${{ vars.AZURE_PROD_REGISTRY_NAME }}
secrets:
aws-role-to-assume: "${{ secrets.PROD_GHA_OIDC_ROLE }}"
docker-hub-username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
docker-hub-password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
secrets: inherit
# This is a bit of a special case so we're not using a generated image map.
add-latest-tag-to-neon-extensions-test-image:
@@ -940,9 +945,7 @@ jobs:
"docker.io/neondatabase/neon-test-extensions-v16:${{ needs.tag.outputs.build-tag }}": ["docker.io/neondatabase/neon-test-extensions-v16:latest"],
"docker.io/neondatabase/neon-test-extensions-v17:${{ needs.tag.outputs.build-tag }}": ["docker.io/neondatabase/neon-test-extensions-v17:latest"]
}
secrets:
docker-hub-username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
docker-hub-password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
secrets: inherit
trigger-custom-extensions-build-and-wait:
needs: [ check-permissions, tag ]

View File

@@ -38,6 +38,9 @@ jobs:
runs-on: us-east-2
container:
image: neondatabase/build-tools:pinned-bookworm
credentials:
username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
options: --init
steps:

View File

@@ -65,6 +65,7 @@ jobs:
permissions:
id-token: write # Required for aws/azure login
packages: write # required for pushing to GHCR
uses: ./.github/workflows/_push-to-container-registry.yml
with:
@@ -72,12 +73,15 @@ jobs:
{
"docker.io/neondatabase/build-tools:${{ inputs.from-tag }}-bullseye": [
"docker.io/neondatabase/build-tools:pinned-bullseye",
"ghcr.io/neondatabase/build-tools:pinned-bullseye",
"${{ vars.NEON_DEV_AWS_ACCOUNT_ID }}.dkr.ecr.${{ vars.AWS_ECR_REGION }}.amazonaws.com/build-tools:pinned-bullseye",
"${{ vars.AZURE_DEV_REGISTRY_NAME }}.azurecr.io/neondatabase/build-tools:pinned-bullseye"
],
"docker.io/neondatabase/build-tools:${{ inputs.from-tag }}-bookworm": [
"docker.io/neondatabase/build-tools:pinned-bookworm",
"docker.io/neondatabase/build-tools:pinned",
"ghcr.io/neondatabase/build-tools:pinned-bookworm",
"ghcr.io/neondatabase/build-tools:pinned",
"${{ vars.NEON_DEV_AWS_ACCOUNT_ID }}.dkr.ecr.${{ vars.AWS_ECR_REGION }}.amazonaws.com/build-tools:pinned-bookworm",
"${{ vars.NEON_DEV_AWS_ACCOUNT_ID }}.dkr.ecr.${{ vars.AWS_ECR_REGION }}.amazonaws.com/build-tools:pinned",
"${{ vars.AZURE_DEV_REGISTRY_NAME }}.azurecr.io/neondatabase/build-tools:pinned-bookworm",
@@ -85,12 +89,10 @@ jobs:
]
}
aws-region: ${{ vars.AWS_ECR_REGION }}
aws-account-ids: "${{ vars.NEON_DEV_AWS_ACCOUNT_ID }}"
aws-account-id: "${{ vars.NEON_DEV_AWS_ACCOUNT_ID }}"
aws-role-to-assume: "gha-oidc-neon-admin"
azure-client-id: ${{ vars.AZURE_DEV_CLIENT_ID }}
azure-subscription-id: ${{ vars.AZURE_DEV_SUBSCRIPTION_ID }}
azure-tenant-id: ${{ vars.AZURE_TENANT_ID }}
acr-registry-name: ${{ vars.AZURE_DEV_REGISTRY_NAME }}
secrets:
aws-role-to-assume: "${{ vars.DEV_AWS_OIDC_ROLE_ARN }}"
docker-hub-username: ${{ secrets.NEON_DOCKERHUB_USERNAME }}
docker-hub-password: ${{ secrets.NEON_DOCKERHUB_PASSWORD }}
secrets: inherit

38
Cargo.lock generated
View File

@@ -1546,6 +1546,17 @@ dependencies = [
"itertools 0.10.5",
]
[[package]]
name = "cron"
version = "0.15.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "5877d3fbf742507b66bc2a1945106bd30dd8504019d596901ddd012a4dd01740"
dependencies = [
"chrono",
"once_cell",
"winnow",
]
[[package]]
name = "crossbeam-channel"
version = "0.5.8"
@@ -1874,6 +1885,12 @@ dependencies = [
"syn 2.0.90",
]
[[package]]
name = "difflib"
version = "0.4.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "6184e33543162437515c2e2b48714794e37845ec9851711914eec9d308f6ebe8"
[[package]]
name = "digest"
version = "0.10.7"
@@ -3331,6 +3348,17 @@ dependencies = [
"wasm-bindgen",
]
[[package]]
name = "json-structural-diff"
version = "0.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "e878e36a8a44c158505c2c818abdc1350413ad83dcb774a0459f6a7ef2b65cbf"
dependencies = [
"difflib",
"regex",
"serde_json",
]
[[package]]
name = "jsonwebtoken"
version = "9.2.0"
@@ -4155,7 +4183,6 @@ dependencies = [
"pageserver_client",
"pageserver_compaction",
"pin-project-lite",
"postgres",
"postgres-protocol",
"postgres-types",
"postgres_backend",
@@ -4242,7 +4269,6 @@ dependencies = [
"futures",
"http-utils",
"pageserver_api",
"postgres",
"reqwest",
"serde",
"thiserror 1.0.69",
@@ -4657,7 +4683,6 @@ dependencies = [
"anyhow",
"itertools 0.10.5",
"once_cell",
"postgres",
"tokio-postgres",
"url",
]
@@ -5799,7 +5824,6 @@ dependencies = [
"once_cell",
"pageserver_api",
"parking_lot 0.12.1",
"postgres",
"postgres-protocol",
"postgres_backend",
"postgres_ffi",
@@ -6433,6 +6457,7 @@ dependencies = [
"chrono",
"clap",
"control_plane",
"cron",
"diesel",
"diesel-async",
"diesel_migrations",
@@ -6443,6 +6468,7 @@ dependencies = [
"humantime",
"hyper 0.14.30",
"itertools 0.10.5",
"json-structural-diff",
"lasso",
"measured",
"metrics",
@@ -8124,9 +8150,9 @@ checksum = "589f6da84c646204747d1270a2a5661ea66ed1cced2631d546fdfb155959f9ec"
[[package]]
name = "winnow"
version = "0.6.13"
version = "0.6.26"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "59b5e5f6c299a3c7890b876a2a587f3115162487e704907d9b6cd29473052ba1"
checksum = "1e90edd2ac1aa278a5c4599b1d89cf03074b610800f866d4026dc199d7929a28"
dependencies = [
"memchr",
]

View File

@@ -77,6 +77,7 @@ byteorder = "1.4"
bytes = "1.9"
camino = "1.1.6"
cfg-if = "1.0.0"
cron = "0.15"
chrono = { version = "0.4", default-features = false, features = ["clock"] }
clap = { version = "4.0", features = ["derive", "env"] }
clashmap = { version = "1.0", features = ["raw-api"] }
@@ -210,6 +211,7 @@ rustls-native-certs = "0.8"
x509-parser = "0.16"
whoami = "1.5.1"
zerocopy = { version = "0.7", features = ["derive"] }
json-structural-diff = { version = "0.2.0" }
## TODO replace this with tracing
env_logger = "0.10"

View File

@@ -395,15 +395,22 @@ RUN case "${PG_VERSION:?}" in \
cd plv8-src && \
if [[ "${PG_VERSION:?}" < "v17" ]]; then patch -p1 < /ext-src/plv8-3.1.10.patch; fi
FROM pg-build AS plv8-build
# Step 1: Build the vendored V8 engine. It doesn't depend on PostgreSQL, so use
# 'build-deps' as the base. This enables caching and avoids unnecessary rebuilds.
# (The V8 engine takes a very long time to build)
FROM build-deps AS plv8-build
ARG PG_VERSION
WORKDIR /ext-src/plv8-src
RUN apt update && \
apt install --no-install-recommends --no-install-suggests -y \
ninja-build python3-dev libncurses5 binutils clang \
&& apt clean && rm -rf /var/lib/apt/lists/*
COPY --from=plv8-src /ext-src/ /ext-src/
WORKDIR /ext-src/plv8-src
RUN make DOCKER=1 -j $(getconf _NPROCESSORS_ONLN) v8
# Step 2: Build the PostgreSQL-dependent parts
COPY --from=pg-build /usr/local/pgsql /usr/local/pgsql
ENV PATH="/usr/local/pgsql/bin:$PATH"
RUN \
# generate and copy upgrade scripts
make generate_upgrades && \
@@ -1451,9 +1458,11 @@ RUN make -j $(getconf _NPROCESSORS_ONLN) && \
FROM build-deps AS pg_mooncake-src
ARG PG_VERSION
WORKDIR /ext-src
COPY compute/patches/duckdb_v113.patch .
RUN wget https://github.com/Mooncake-Labs/pg_mooncake/releases/download/v0.1.2/pg_mooncake-0.1.2.tar.gz -O pg_mooncake.tar.gz && \
echo "4550473784fcdd2e1e18062bc01eb9c286abd27cdf5e11a4399be6c0a426ba90 pg_mooncake.tar.gz" | sha256sum --check && \
mkdir pg_mooncake-src && cd pg_mooncake-src && tar xzf ../pg_mooncake.tar.gz --strip-components=1 -C . && \
cd third_party/duckdb && patch -p1 < /ext-src/duckdb_v113.patch && cd ../.. && \
echo "make -f pg_mooncake-src/Makefile.build installcheck TEST_DIR=./test SQL_DIR=./sql SRC_DIR=./src" > neon-test.sh && \
chmod a+x neon-test.sh
@@ -1473,6 +1482,7 @@ RUN make release -j $(getconf _NPROCESSORS_ONLN) && \
FROM build-deps AS pg_duckdb-src
WORKDIR /ext-src
COPY compute/patches/pg_duckdb_v031.patch .
COPY compute/patches/duckdb_v120.patch .
# pg_duckdb build requires source dir to be a git repo to get submodules
# allow neon_superuser to execute some functions that in pg_duckdb are available to superuser only:
# - extension management function duckdb.install_extension()
@@ -1480,7 +1490,9 @@ COPY compute/patches/pg_duckdb_v031.patch .
RUN git clone --depth 1 --branch v0.3.1 https://github.com/duckdb/pg_duckdb.git pg_duckdb-src && \
cd pg_duckdb-src && \
git submodule update --init --recursive && \
patch -p1 < /ext-src/pg_duckdb_v031.patch
patch -p1 < /ext-src/pg_duckdb_v031.patch && \
cd third_party/duckdb && \
patch -p1 < /ext-src/duckdb_v120.patch
FROM pg-build AS pg_duckdb-build
ARG PG_VERSION
@@ -1669,11 +1681,7 @@ COPY --from=pg_anon-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_ivm-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_partman-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_mooncake-build /usr/local/pgsql/ /usr/local/pgsql/
# Disabled temporarily, because it clashed with pg_mooncake. pg_mooncake
# also depends on libduckdb, but a different version.
#COPY --from=pg_duckdb-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_duckdb-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pg_repack-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pgaudit-build /usr/local/pgsql/ /usr/local/pgsql/
COPY --from=pgauditlogtofile-build /usr/local/pgsql/ /usr/local/pgsql/
@@ -1810,7 +1818,7 @@ RUN make PG_VERSION="${PG_VERSION:?}" -C compute
FROM pg-build AS extension-tests
ARG PG_VERSION
RUN mkdir /ext-src
COPY docker-compose/ext-src/ /ext-src/
COPY --from=pg-build /postgres /postgres
#COPY --from=postgis-src /ext-src/ /ext-src/

View File

@@ -0,0 +1,25 @@
diff --git a/libduckdb.map b/libduckdb.map
new file mode 100644
index 0000000000..3b56f00cd7
--- /dev/null
+++ b/libduckdb.map
@@ -0,0 +1,6 @@
+DUCKDB_1.1.3 {
+ global:
+ *duckdb*;
+ local:
+ *;
+};
diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index 3e757a4bcc..88ab4005b9 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -135,6 +135,8 @@ else()
target_link_libraries(duckdb ${DUCKDB_LINK_LIBS})
link_threads(duckdb)
link_extension_libraries(duckdb)
+ target_link_options(duckdb PRIVATE
+ -Wl,--version-script=${CMAKE_SOURCE_DIR}/libduckdb.map)
add_library(duckdb_static STATIC ${ALL_OBJECT_FILES})
target_link_libraries(duckdb_static ${DUCKDB_LINK_LIBS})

View File

@@ -0,0 +1,67 @@
diff --git a/libduckdb_pg_duckdb.map b/libduckdb_pg_duckdb.map
new file mode 100644
index 0000000000..0872978b48
--- /dev/null
+++ b/libduckdb_pg_duckdb.map
@@ -0,0 +1,6 @@
+DUCKDB_1.2.0 {
+ global:
+ *duckdb*;
+ local:
+ *;
+};
diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index 58adef3fc0..2c522f91be 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -59,7 +59,7 @@ endfunction()
if(AMALGAMATION_BUILD)
- add_library(duckdb SHARED "${PROJECT_SOURCE_DIR}/src/amalgamation/duckdb.cpp")
+ add_library(duckdb_pg_duckdb SHARED "${PROJECT_SOURCE_DIR}/src/amalgamation/duckdb.cpp")
target_link_libraries(duckdb ${DUCKDB_SYSTEM_LIBS})
link_threads(duckdb)
link_extension_libraries(duckdb)
@@ -109,7 +109,7 @@ else()
duckdb_yyjson
duckdb_zstd)
- add_library(duckdb SHARED ${ALL_OBJECT_FILES})
+ add_library(duckdb_pg_duckdb SHARED ${ALL_OBJECT_FILES})
if(WIN32 AND NOT MINGW)
ensure_variable_is_number(DUCKDB_MAJOR_VERSION RC_MAJOR_VERSION)
@@ -131,9 +131,11 @@ else()
target_sources(duckdb PRIVATE version.rc)
endif()
- target_link_libraries(duckdb ${DUCKDB_LINK_LIBS})
- link_threads(duckdb)
- link_extension_libraries(duckdb)
+ target_link_libraries(duckdb_pg_duckdb ${DUCKDB_LINK_LIBS})
+ link_threads(duckdb_pg_duckdb)
+ link_extension_libraries(duckdb_pg_duckdb)
+ target_link_options(duckdb_pg_duckdb PRIVATE
+ -Wl,--version-script=${CMAKE_SOURCE_DIR}/libduckdb_pg_duckdb.map)
add_library(duckdb_static STATIC ${ALL_OBJECT_FILES})
target_link_libraries(duckdb_static ${DUCKDB_LINK_LIBS})
@@ -141,7 +143,7 @@ else()
link_extension_libraries(duckdb_static)
target_include_directories(
- duckdb PUBLIC $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/include>
+ duckdb_pg_duckdb PUBLIC $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/include>
$<INSTALL_INTERFACE:${CMAKE_INSTALL_INCLUDEDIR}>)
target_include_directories(
@@ -161,7 +163,7 @@ else()
endif()
install(
- TARGETS duckdb duckdb_static
+ TARGETS duckdb_pg_duckdb duckdb_static
EXPORT "${DUCKDB_EXPORT_SET}"
LIBRARY DESTINATION "${INSTALL_LIB_DIR}"
ARCHIVE DESTINATION "${INSTALL_LIB_DIR}"

View File

@@ -1,3 +1,25 @@
diff --git a/Makefile b/Makefile
index 3235cc8..6b892bc 100644
--- a/Makefile
+++ b/Makefile
@@ -32,7 +32,7 @@ else
DUCKDB_BUILD_TYPE = release
endif
-DUCKDB_LIB = libduckdb$(DLSUFFIX)
+DUCKDB_LIB = libduckdb_pg_duckdb$(DLSUFFIX)
FULL_DUCKDB_LIB = third_party/duckdb/build/$(DUCKDB_BUILD_TYPE)/src/$(DUCKDB_LIB)
ERROR_ON_WARNING ?=
@@ -54,7 +54,7 @@ override PG_CXXFLAGS += -std=c++17 ${DUCKDB_BUILD_CXX_FLAGS} ${COMPILER_FLAGS} -
# changes to the vendored code in one place.
override PG_CFLAGS += -Wno-declaration-after-statement
-SHLIB_LINK += -Wl,-rpath,$(PG_LIB)/ -lpq -Lthird_party/duckdb/build/$(DUCKDB_BUILD_TYPE)/src -L$(PG_LIB) -lduckdb -lstdc++ -llz4
+SHLIB_LINK += -Wl,-rpath,$(PG_LIB)/ -lpq -Lthird_party/duckdb/build/$(DUCKDB_BUILD_TYPE)/src -L$(PG_LIB) -lduckdb_pg_duckdb -lstdc++ -llz4
include Makefile.global
diff --git a/sql/pg_duckdb--0.2.0--0.3.0.sql b/sql/pg_duckdb--0.2.0--0.3.0.sql
index d777d76..af60106 100644
--- a/sql/pg_duckdb--0.2.0--0.3.0.sql

View File

@@ -44,6 +44,11 @@ shutdownHook: |
files:
- filename: compute_ctl-sudoers
content: |
# Reverse hostname lookup doesn't currently work, and isn't needed anyway when all
# the rules use ALL as the hostname. Avoid the pointless lookups and the "unable to
# resolve host" log messages that they generate.
Defaults !fqdn
# Allow postgres user (which is what compute_ctl runs as) to run /neonvm/bin/resize-swap
# and /neonvm/bin/set-disk-quota as root without requiring entering a password (NOPASSWD),
# regardless of hostname (ALL)

View File

@@ -44,6 +44,11 @@ shutdownHook: |
files:
- filename: compute_ctl-sudoers
content: |
# Reverse hostname lookup doesn't currently work, and isn't needed anyway when all
# the rules use ALL as the hostname. Avoid the pointless lookups and the "unable to
# resolve host" log messages that they generate.
Defaults !fqdn
# Allow postgres user (which is what compute_ctl runs as) to run /neonvm/bin/resize-swap
# and /neonvm/bin/set-disk-quota as root without requiring entering a password (NOPASSWD),
# regardless of hostname (ALL)

View File

@@ -1,7 +1,7 @@
[package]
name = "compute_tools"
version = "0.1.0"
edition.workspace = true
edition = "2024"
license.workspace = true
[features]

View File

@@ -40,34 +40,31 @@ use std::path::Path;
use std::process::exit;
use std::str::FromStr;
use std::sync::atomic::Ordering;
use std::sync::{mpsc, Arc, Condvar, Mutex, RwLock};
use std::{thread, time::Duration};
use std::sync::{Arc, Condvar, Mutex, RwLock, mpsc};
use std::thread;
use std::time::Duration;
use anyhow::{Context, Result};
use chrono::Utc;
use clap::Parser;
use compute_tools::disk_quota::set_disk_quota;
use compute_tools::http::server::Server;
use compute_tools::lsn_lease::launch_lsn_lease_bg_task_for_static;
use signal_hook::consts::{SIGQUIT, SIGTERM};
use signal_hook::{consts::SIGINT, iterator::Signals};
use tracing::{error, info, warn};
use url::Url;
use compute_api::responses::{ComputeCtlConfig, ComputeStatus};
use compute_api::spec::ComputeSpec;
use compute_tools::compute::{
forward_termination_signal, ComputeNode, ComputeState, ParsedSpec, PG_PID,
ComputeNode, ComputeState, PG_PID, ParsedSpec, forward_termination_signal,
};
use compute_tools::configurator::launch_configurator;
use compute_tools::extension_server::get_pg_version_string;
use compute_tools::http::server::Server;
use compute_tools::logger::*;
use compute_tools::lsn_lease::launch_lsn_lease_bg_task_for_static;
use compute_tools::monitor::launch_monitor;
use compute_tools::params::*;
use compute_tools::spec::*;
use compute_tools::swap::resize_swap;
use rlimit::{setrlimit, Resource};
use rlimit::{Resource, setrlimit};
use signal_hook::consts::{SIGINT, SIGQUIT, SIGTERM};
use signal_hook::iterator::Signals;
use tracing::{error, info};
use url::Url;
use utils::failpoint_support;
// this is an arbitrary build tag. Fine as a default / for testing purposes
@@ -149,6 +146,8 @@ struct Cli {
fn main() -> Result<()> {
let cli = Cli::parse();
let scenario = failpoint_support::init();
// For historical reasons, the main thread that processes the spec and launches postgres
// is synchronous, but we always have this tokio runtime available and we "enter" it so
// that you can use tokio::spawn() and tokio::runtime::Handle::current().block_on(...)
@@ -160,8 +159,6 @@ fn main() -> Result<()> {
let build_tag = runtime.block_on(init())?;
let scenario = failpoint_support::init();
// enable core dumping for all child processes
setrlimit(Resource::CORE, rlimit::INFINITY, rlimit::INFINITY)?;
@@ -173,7 +170,11 @@ fn main() -> Result<()> {
let compute = wait_spec(build_tag, &cli, cli_spec)?;
start_postgres(&cli, compute)?
bootstrap_compute(
compute,
#[cfg(target_os = "linux")]
&cli,
)
// Startup is finished, exit the startup tracing span
};
@@ -349,6 +350,8 @@ fn wait_spec(
ext_remote_storage: cli.remote_ext_config.clone(),
ext_download_progress: RwLock::new(HashMap::new()),
build_tag,
resize_swap_on_bind: cli.resize_swap_on_bind,
set_disk_quota_for_fs: cli.set_disk_quota_for_fs.clone(),
};
let compute = Arc::new(compute_node);
@@ -401,10 +404,11 @@ fn wait_spec(
Ok(compute)
}
fn start_postgres(
cli: &Cli,
// Start Postgres and some aux threads like various monitors
fn bootstrap_compute(
compute: Arc<ComputeNode>,
) -> Result<(Option<PostgresHandle>, StartPostgresResult)> {
#[cfg(target_os = "linux")] cli: &Cli,
) -> (Option<PostgresHandle>, StartPostgresResult) {
// We got all we need, update the state.
let mut state = compute.state.lock().unwrap();
state.set_status(ComputeStatus::Init, &compute.state_changed);
@@ -413,10 +417,9 @@ fn start_postgres(
"running compute with features: {:?}",
state.pspec.as_ref().unwrap().spec.features
);
// before we release the mutex, fetch some parameters for later.
// Before we release the mutex, fetch some parameters for later.
let &ComputeSpec {
swap_size_bytes,
disk_quota_bytes,
#[cfg(target_os = "linux")]
disable_lfc_resizing,
..
@@ -427,120 +430,86 @@ fn start_postgres(
let _monitor_handle = launch_monitor(&compute);
let _configurator_handle = launch_configurator(&compute);
let mut prestartup_failed = false;
let mut delay_exit = false;
// Resize swap to the desired size if the compute spec says so
if let (Some(size_bytes), true) = (swap_size_bytes, cli.resize_swap_on_bind) {
// To avoid 'swapoff' hitting postgres startup, we need to run resize-swap to completion
// *before* starting postgres.
//
// In theory, we could do this asynchronously if SkipSwapon was enabled for VMs, but this
// carries a risk of introducing hard-to-debug issues - e.g. if postgres sometimes gets
// OOM-killed during startup because swap wasn't available yet.
match resize_swap(size_bytes) {
Ok(()) => {
let size_mib = size_bytes as f32 / (1 << 20) as f32; // just for more coherent display.
info!(%size_bytes, %size_mib, "resized swap");
}
Err(err) => {
let err = err.context("failed to resize swap");
error!("{err:#}");
// Mark compute startup as failed; don't try to start postgres, and report this
// error to the control plane when it next asks.
prestartup_failed = true;
compute.set_failed_status(err);
delay_exit = true;
}
}
}
// Set disk quota if the compute spec says so
if let (Some(disk_quota_bytes), Some(disk_quota_fs_mountpoint)) =
(disk_quota_bytes, cli.set_disk_quota_for_fs.as_ref())
{
match set_disk_quota(disk_quota_bytes, disk_quota_fs_mountpoint) {
Ok(()) => {
let size_mib = disk_quota_bytes as f32 / (1 << 20) as f32; // just for more coherent display.
info!(%disk_quota_bytes, %size_mib, "set disk quota");
}
Err(err) => {
let err = err.context("failed to set disk quota");
error!("{err:#}");
// Mark compute startup as failed; don't try to start postgres, and report this
// error to the control plane when it next asks.
prestartup_failed = true;
compute.set_failed_status(err);
delay_exit = true;
}
}
}
// Start Postgres
let mut pg = None;
if !prestartup_failed {
pg = match compute.start_compute() {
Ok(pg) => {
info!(postmaster_pid = %pg.0.id(), "Postgres was started");
Some(pg)
match compute.start_compute() {
Ok(pg) => {
info!(postmaster_pid = %pg.0.id(), "Postgres was started");
// Start the vm-monitor if directed to. The vm-monitor only runs on linux
// because it requires cgroups.
cfg_if::cfg_if! {
if #[cfg(target_os = "linux")] {
use std::env;
use tokio_util::sync::CancellationToken;
// This token is used internally by the monitor to clean up all threads
let token = CancellationToken::new();
// Don't pass postgres connection string to vm-monitor if we don't want it to resize LFC
let pgconnstr = if disable_lfc_resizing.unwrap_or(false) {
None
} else {
Some(cli.filecache_connstr.clone())
};
let vm_monitor = if env::var_os("AUTOSCALING").is_some() {
let vm_monitor = tokio::spawn(vm_monitor::start(
Box::leak(Box::new(vm_monitor::Args {
cgroup: Some(cli.cgroup.clone()),
pgconnstr,
addr: cli.vm_monitor_addr.clone(),
})),
token.clone(),
));
Some(vm_monitor)
} else {
None
};
}
}
Err(err) => {
error!("could not start the compute node: {:#}", err);
compute.set_failed_status(err);
delay_exit = true;
None
(
Some(pg),
StartPostgresResult {
delay_exit,
compute,
#[cfg(target_os = "linux")]
token,
#[cfg(target_os = "linux")]
vm_monitor,
},
)
}
Err(err) => {
error!("could not start the compute node: {:#}", err);
compute.set_failed_status(err);
delay_exit = true;
cfg_if::cfg_if! {
if #[cfg(target_os = "linux")] {
use tokio_util::sync::CancellationToken;
let token = CancellationToken::new();
(
None,
StartPostgresResult {
delay_exit,
compute,
token,
vm_monitor: None,
},
)
} else {
(None, StartPostgresResult {
delay_exit,
compute,
})
}
}
};
} else {
warn!("skipping postgres startup because pre-startup step failed");
}
// Start the vm-monitor if directed to. The vm-monitor only runs on linux
// because it requires cgroups.
cfg_if::cfg_if! {
if #[cfg(target_os = "linux")] {
use std::env;
use tokio_util::sync::CancellationToken;
// This token is used internally by the monitor to clean up all threads
let token = CancellationToken::new();
// don't pass postgres connection string to vm-monitor if we don't want it to resize LFC
let pgconnstr = if disable_lfc_resizing.unwrap_or(false) {
None
} else {
Some(cli.filecache_connstr.clone())
};
let vm_monitor = if env::var_os("AUTOSCALING").is_some() {
let vm_monitor = tokio::spawn(vm_monitor::start(
Box::leak(Box::new(vm_monitor::Args {
cgroup: Some(cli.cgroup.clone()),
pgconnstr,
addr: cli.vm_monitor_addr.clone(),
})),
token.clone(),
));
Some(vm_monitor)
} else {
None
};
}
}
Ok((
pg,
StartPostgresResult {
delay_exit,
compute,
#[cfg(target_os = "linux")]
token,
#[cfg(target_os = "linux")]
vm_monitor,
},
))
}
type PostgresHandle = (std::process::Child, tokio::task::JoinHandle<Result<()>>);

View File

@@ -25,13 +25,13 @@
//! docker push localhost:3030/localregistry/compute-node-v14:latest
//! ```
use anyhow::{bail, Context};
use anyhow::{Context, bail};
use aws_config::BehaviorVersion;
use camino::{Utf8Path, Utf8PathBuf};
use clap::{Parser, Subcommand};
use compute_tools::extension_server::{get_pg_version, PostgresMajorVersion};
use compute_tools::extension_server::{PostgresMajorVersion, get_pg_version};
use nix::unistd::Pid;
use tracing::{error, info, info_span, warn, Instrument};
use tracing::{Instrument, error, info, info_span, warn};
use utils::fs_ext::is_directory_empty;
#[path = "fast_import/aws_s3_sync.rs"]
@@ -558,7 +558,9 @@ async fn cmd_dumprestore(
decode_connstring(kms_client.as_ref().unwrap(), &key_id, dest_ciphertext)
.await?
} else {
bail!("destination connection string must be provided in spec for dump_restore command");
bail!(
"destination connection string must be provided in spec for dump_restore command"
);
};
(source, dest)

View File

@@ -1,11 +1,10 @@
use camino::{Utf8Path, Utf8PathBuf};
use tokio::task::JoinSet;
use tracing::{info, warn};
use walkdir::WalkDir;
use super::s3_uri::S3Uri;
use tracing::{info, warn};
const MAX_PARALLEL_UPLOADS: usize = 10;
/// Upload all files from 'local' to 'remote'

View File

@@ -1,6 +1,7 @@
use anyhow::Result;
use std::str::FromStr;
use anyhow::Result;
/// Struct to hold parsed S3 components
#[derive(Debug, Clone, PartialEq, Eq)]
pub struct S3Uri {

View File

@@ -1,18 +1,20 @@
use std::path::Path;
use std::process::Stdio;
use std::result::Result;
use std::sync::Arc;
use compute_api::responses::CatalogObjects;
use futures::Stream;
use postgres::NoTls;
use std::{path::Path, process::Stdio, result::Result, sync::Arc};
use tokio::{
io::{AsyncBufReadExt, BufReader},
process::Command,
spawn,
};
use tokio::io::{AsyncBufReadExt, BufReader};
use tokio::process::Command;
use tokio::spawn;
use tokio_stream::{self as stream, StreamExt};
use tokio_util::codec::{BytesCodec, FramedRead};
use tracing::warn;
use crate::compute::ComputeNode;
use crate::pg_helpers::{get_existing_dbs_async, get_existing_roles_async, postgres_conf_for_db};
use compute_api::responses::CatalogObjects;
pub async fn get_dbs_and_roles(compute: &Arc<ComputeNode>) -> anyhow::Result<CatalogObjects> {
let conf = compute.get_tokio_conn_conf(Some("compute_ctl:get_dbs_and_roles"));
@@ -55,7 +57,7 @@ pub enum SchemaDumpError {
pub async fn get_database_schema(
compute: &Arc<ComputeNode>,
dbname: &str,
) -> Result<impl Stream<Item = Result<bytes::Bytes, std::io::Error>>, SchemaDumpError> {
) -> Result<impl Stream<Item = Result<bytes::Bytes, std::io::Error>> + use<>, SchemaDumpError> {
let pgbin = &compute.pgbin;
let basepath = Path::new(pgbin).parent().unwrap();
let pgdump = basepath.join("pg_dump");

View File

@@ -1,4 +1,4 @@
use anyhow::{anyhow, Ok, Result};
use anyhow::{Ok, Result, anyhow};
use tokio_postgres::NoTls;
use tracing::{error, instrument, warn};

View File

@@ -1,42 +1,38 @@
use std::collections::{HashMap, HashSet};
use std::env;
use std::fs;
use std::iter::once;
use std::os::unix::fs::{symlink, PermissionsExt};
use std::os::unix::fs::{PermissionsExt, symlink};
use std::path::Path;
use std::process::{Command, Stdio};
use std::str::FromStr;
use std::sync::atomic::AtomicU32;
use std::sync::atomic::Ordering;
use std::sync::atomic::{AtomicU32, Ordering};
use std::sync::{Arc, Condvar, Mutex, RwLock};
use std::time::Duration;
use std::time::Instant;
use std::time::{Duration, Instant};
use std::{env, fs};
use anyhow::{Context, Result};
use chrono::{DateTime, Utc};
use compute_api::spec::{Database, PgIdent, Role};
use compute_api::privilege::Privilege;
use compute_api::responses::{ComputeMetrics, ComputeStatus};
use compute_api::spec::{
ComputeFeature, ComputeMode, ComputeSpec, Database, ExtVersion, PgIdent, Role,
};
use futures::StreamExt;
use futures::future::join_all;
use futures::stream::FuturesUnordered;
use futures::StreamExt;
use nix::sys::signal::{Signal, kill};
use nix::unistd::Pid;
use postgres;
use postgres::error::SqlState;
use postgres::NoTls;
use postgres::error::SqlState;
use remote_storage::{DownloadError, RemotePath};
use tokio::spawn;
use tracing::{debug, error, info, instrument, warn};
use utils::id::{TenantId, TimelineId};
use utils::lsn::Lsn;
use compute_api::privilege::Privilege;
use compute_api::responses::{ComputeMetrics, ComputeStatus};
use compute_api::spec::{ComputeFeature, ComputeMode, ComputeSpec, ExtVersion};
use utils::measured_stream::MeasuredReader;
use nix::sys::signal::{kill, Signal};
use remote_storage::{DownloadError, RemotePath};
use tokio::spawn;
use crate::disk_quota::set_disk_quota;
use crate::installed_extensions::get_installed_extensions;
use crate::local_proxy;
use crate::pg_helpers::*;
use crate::spec::*;
use crate::spec_apply::ApplySpecPhase::{
@@ -45,14 +41,13 @@ use crate::spec_apply::ApplySpecPhase::{
HandleNeonExtension, HandleOtherExtensions, RenameAndDeleteDatabases, RenameRoles,
RunInEachDatabase,
};
use crate::spec_apply::PerDatabasePhase;
use crate::spec_apply::PerDatabasePhase::{
ChangeSchemaPerms, DeleteDBRoleReferences, DropLogicalSubscriptions, HandleAnonExtension,
};
use crate::spec_apply::{apply_operations, MutableApplyContext, DB};
use crate::spec_apply::{DB, MutableApplyContext, PerDatabasePhase, apply_operations};
use crate::swap::resize_swap;
use crate::sync_sk::{check_if_synced, ping_safekeeper};
use crate::{config, extension_server};
use crate::{config, extension_server, local_proxy};
pub static SYNC_SAFEKEEPERS_PID: AtomicU32 = AtomicU32::new(0);
pub static PG_PID: AtomicU32 = AtomicU32::new(0);
@@ -98,6 +93,10 @@ pub struct ComputeNode {
// key: ext_archive_name, value: started download time, download_completed?
pub ext_download_progress: RwLock<HashMap<String, (DateTime<Utc>, bool)>>,
pub build_tag: String,
/// Initialized from cli.resize_swap_on_bind
pub resize_swap_on_bind: bool,
/// Initialized from cli.set_disk_quota_for_fs
pub set_disk_quota_for_fs: Option<String>,
}
// store some metrics about download size that might impact startup time
@@ -546,6 +545,7 @@ impl ComputeNode {
// Fast path for sync_safekeepers. If they're already synced we get the lsn
// in one roundtrip. If not, we should do a full sync_safekeepers.
#[instrument(skip_all)]
pub fn check_safekeepers_synced(&self, compute_state: &ComputeState) -> Result<Option<Lsn>> {
let start_time = Utc::now();
@@ -1317,7 +1317,7 @@ impl ComputeNode {
// Merge-apply spec & changes to PostgreSQL state.
self.apply_spec_sql(spec.clone(), conf.clone(), max_concurrent_connections)?;
if let Some(ref local_proxy) = &spec.clone().local_proxy_config {
if let Some(local_proxy) = &spec.clone().local_proxy_config {
info!("configuring local_proxy");
local_proxy::configure(local_proxy).context("apply_config local_proxy")?;
}
@@ -1434,12 +1434,59 @@ impl ComputeNode {
Ok(())
}
/// Configure some VM parameters like swap and disk quota
#[instrument(skip_all)]
pub fn configure_vm(&self, spec: &ComputeSpec) -> Result<()> {
// Resize swap to the desired size if the compute spec says so
if let (Some(size_bytes), true) = (spec.swap_size_bytes, self.resize_swap_on_bind) {
// To avoid 'swapoff' hitting postgres startup, we need to run resize-swap to completion
// *before* starting postgres.
//
// In theory, we could do this asynchronously if SkipSwapon was enabled for VMs, but this
// carries a risk of introducing hard-to-debug issues - e.g. if postgres sometimes gets
// OOM-killed during startup because swap wasn't available yet.
match resize_swap(size_bytes) {
Ok(()) => {
let size_mib = size_bytes as f32 / (1 << 20) as f32; // just for more coherent display.
info!(%size_bytes, %size_mib, "resized swap");
}
Err(err) => {
let err = err.context("failed to resize swap");
error!("{err:#}");
return Err(err);
}
}
}
// Set disk quota if the compute spec says so
if let (Some(disk_quota_bytes), Some(disk_quota_fs_mountpoint)) =
(spec.disk_quota_bytes, self.set_disk_quota_for_fs.as_ref())
{
match set_disk_quota(disk_quota_bytes, disk_quota_fs_mountpoint) {
Ok(()) => {
let size_mib = disk_quota_bytes as f32 / (1 << 20) as f32; // just for more coherent display.
info!(%disk_quota_bytes, %size_mib, "set disk quota");
}
Err(err) => {
let err = err.context("failed to set disk quota");
error!("{err:#}");
return Err(err);
}
}
}
Ok(())
}
#[instrument(skip_all)]
pub fn start_compute(
&self,
) -> Result<(std::process::Child, tokio::task::JoinHandle<Result<()>>)> {
let compute_state = self.state.lock().unwrap().clone();
let pspec = compute_state.pspec.as_ref().expect("spec must be set");
info!(
"starting compute for project {}, operation {}, tenant {}, timeline {}",
pspec.spec.cluster.cluster_id.as_deref().unwrap_or("None"),
@@ -1448,9 +1495,11 @@ impl ComputeNode {
pspec.timeline_id,
);
// tune pgbouncer
self.configure_vm(&pspec.spec)?;
// Configure pgbouncer
if let Some(pgbouncer_settings) = &pspec.spec.pgbouncer_settings {
info!("tuning pgbouncer");
info!("configuring pgbouncer");
// Spawn a background task to do the tuning,
// so that we don't block the main thread that starts Postgres.
@@ -1476,14 +1525,11 @@ impl ComputeNode {
});
}
info!(
"start_compute spec.remote_extensions {:?}",
pspec.spec.remote_extensions
);
// This part is sync, because we need to download
// remote shared_preload_libraries before postgres start (if any)
if let Some(remote_extensions) = &pspec.spec.remote_extensions {
info!(?remote_extensions, "processing remote extensions");
// First, create control files for all availale extensions
extension_server::create_control_files(remote_extensions, &self.pgbin);
@@ -1502,7 +1548,7 @@ impl ComputeNode {
state.metrics.largest_ext_size = remote_ext_metrics.largest_ext_size;
state.metrics.total_ext_download_size = remote_ext_metrics.total_ext_download_size;
info!(
"Loading shared_preload_libraries took {:?}ms",
"loading shared_preload_libraries from remote extensions took {:?}ms",
library_load_time
);
info!("{:?}", remote_ext_metrics);
@@ -1537,7 +1583,9 @@ impl ComputeNode {
&postgresql_conf_path,
"neon.disable_logical_replication_subscribers=false",
)? {
info!("updated postgresql.conf to set neon.disable_logical_replication_subscribers=false");
info!(
"updated postgresql.conf to set neon.disable_logical_replication_subscribers=false"
);
}
self.pg_reload_conf()?;
}
@@ -1559,6 +1607,7 @@ impl ComputeNode {
});
}
let metrics: ComputeMetrics;
let startup_end_time = Utc::now();
{
let mut state = self.state.lock().unwrap();
@@ -1577,21 +1626,17 @@ impl ComputeNode {
.to_std()
.unwrap()
.as_millis() as u64;
metrics = state.metrics.clone();
}
self.set_status(ComputeStatus::Running);
// Log metrics so that we can search for slow operations in logs
info!(
?metrics,
"finished configuration of compute for project {}",
pspec.spec.cluster.cluster_id.as_deref().unwrap_or("None")
);
// Log metrics so that we can search for slow operations in logs
let metrics = {
let state = self.state.lock().unwrap();
state.metrics.clone()
};
info!(?metrics, "compute start finished");
Ok(pg_process)
}
@@ -1764,7 +1809,9 @@ LIMIT 100",
info!("extension already downloaded, skipping re-download");
return Ok(0);
} else if start_time_delta < HANG_TIMEOUT && !first_try {
info!("download {ext_archive_name} already started by another process, hanging untill completion or timeout");
info!(
"download {ext_archive_name} already started by another process, hanging untill completion or timeout"
);
let mut interval = tokio::time::interval(tokio::time::Duration::from_millis(500));
loop {
info!("waiting for download");

View File

@@ -4,11 +4,10 @@ use std::io::prelude::*;
use std::path::Path;
use anyhow::Result;
use crate::pg_helpers::escape_conf_value;
use crate::pg_helpers::{GenericOptionExt, PgOptionsSerialize};
use compute_api::spec::{ComputeMode, ComputeSpec, GenericOption};
use crate::pg_helpers::{GenericOptionExt, PgOptionsSerialize, escape_conf_value};
/// Check that `line` is inside a text file and put it there if it is not.
/// Create file if it doesn't exist.
pub fn line_in_file(path: &Path, line: &str) -> Result<bool> {

View File

@@ -1,9 +1,8 @@
use std::sync::Arc;
use std::thread;
use tracing::{error, info, instrument};
use compute_api::responses::ComputeStatus;
use tracing::{error, info, instrument};
use crate::compute::ComputeNode;

View File

@@ -1,9 +1,11 @@
use anyhow::Context;
use tracing::instrument;
pub const DISK_QUOTA_BIN: &str = "/neonvm/bin/set-disk-quota";
/// If size_bytes is 0, it disables the quota. Otherwise, it sets filesystem quota to size_bytes.
/// `fs_mountpoint` should point to the mountpoint of the filesystem where the quota should be set.
#[instrument]
pub fn set_disk_quota(size_bytes: u64, fs_mountpoint: &str) -> anyhow::Result<()> {
let size_kb = size_bytes / 1024;
// run `/neonvm/bin/set-disk-quota {size_kb} {mountpoint}`

View File

@@ -71,15 +71,15 @@ More specifically, here is an example ext_index.json
}
}
*/
use anyhow::Result;
use anyhow::{bail, Context};
use std::path::Path;
use std::str;
use anyhow::{Context, Result, bail};
use bytes::Bytes;
use compute_api::spec::RemoteExtSpec;
use regex::Regex;
use remote_storage::*;
use reqwest::StatusCode;
use std::path::Path;
use std::str;
use tar::Archive;
use tracing::info;
use tracing::log::warn;
@@ -244,7 +244,10 @@ pub fn create_control_files(remote_extensions: &RemoteExtSpec, pgbin: &str) {
info!("writing file {:?}{:?}", control_path, control_content);
std::fs::write(control_path, control_content).unwrap();
} else {
warn!("control file {:?} exists both locally and remotely. ignoring the remote version.", control_path);
warn!(
"control file {:?} exists both locally and remotely. ignoring the remote version.",
control_path
);
}
}
}

View File

@@ -1,6 +1,7 @@
use std::ops::{Deref, DerefMut};
use axum::extract::{rejection::JsonRejection, FromRequest, Request};
use axum::extract::rejection::JsonRejection;
use axum::extract::{FromRequest, Request};
use compute_api::responses::GenericAPIError;
use http::StatusCode;

View File

@@ -1,8 +1,10 @@
use std::ops::{Deref, DerefMut};
use axum::extract::{rejection::PathRejection, FromRequestParts};
use axum::extract::FromRequestParts;
use axum::extract::rejection::PathRejection;
use compute_api::responses::GenericAPIError;
use http::{request::Parts, StatusCode};
use http::StatusCode;
use http::request::Parts;
/// Custom `Path` extractor, so that we can format errors into
/// `JsonResponse<GenericAPIError>`.

View File

@@ -1,8 +1,10 @@
use std::ops::{Deref, DerefMut};
use axum::extract::{rejection::QueryRejection, FromRequestParts};
use axum::extract::FromRequestParts;
use axum::extract::rejection::QueryRejection;
use compute_api::responses::GenericAPIError;
use http::{request::Parts, StatusCode};
use http::StatusCode;
use http::request::Parts;
/// Custom `Query` extractor, so that we can format errors into
/// `JsonResponse<GenericAPIError>`.

View File

@@ -1,6 +1,8 @@
use axum::{body::Body, response::Response};
use axum::body::Body;
use axum::response::Response;
use compute_api::responses::{ComputeStatus, GenericAPIError};
use http::{header::CONTENT_TYPE, StatusCode};
use http::StatusCode;
use http::header::CONTENT_TYPE;
use serde::Serialize;
use tracing::error;

View File

@@ -1,10 +1,13 @@
use std::sync::Arc;
use axum::{extract::State, response::Response};
use axum::extract::State;
use axum::response::Response;
use compute_api::responses::ComputeStatus;
use http::StatusCode;
use crate::{checker::check_writability, compute::ComputeNode, http::JsonResponse};
use crate::checker::check_writability;
use crate::compute::ComputeNode;
use crate::http::JsonResponse;
/// Check that the compute is currently running.
pub(in crate::http) async fn is_writable(State(compute): State<Arc<ComputeNode>>) -> Response {

View File

@@ -1,18 +1,16 @@
use std::sync::Arc;
use axum::{extract::State, response::Response};
use compute_api::{
requests::ConfigurationRequest,
responses::{ComputeStatus, ComputeStatusResponse},
};
use axum::extract::State;
use axum::response::Response;
use compute_api::requests::ConfigurationRequest;
use compute_api::responses::{ComputeStatus, ComputeStatusResponse};
use http::StatusCode;
use tokio::task;
use tracing::info;
use crate::{
compute::{ComputeNode, ParsedSpec},
http::{extract::Json, JsonResponse},
};
use crate::compute::{ComputeNode, ParsedSpec};
use crate::http::JsonResponse;
use crate::http::extract::Json;
// Accept spec in JSON format and request compute configuration. If anything
// goes wrong after we set the compute status to `ConfigurationPending` and

View File

@@ -1,14 +1,16 @@
use std::sync::Arc;
use axum::{body::Body, extract::State, response::Response};
use http::{header::CONTENT_TYPE, StatusCode};
use axum::body::Body;
use axum::extract::State;
use axum::response::Response;
use http::StatusCode;
use http::header::CONTENT_TYPE;
use serde::Deserialize;
use crate::{
catalog::{get_database_schema, SchemaDumpError},
compute::ComputeNode,
http::{extract::Query, JsonResponse},
};
use crate::catalog::{SchemaDumpError, get_database_schema};
use crate::compute::ComputeNode;
use crate::http::JsonResponse;
use crate::http::extract::Query;
#[derive(Debug, Clone, Deserialize)]
pub(in crate::http) struct DatabaseSchemaParams {

View File

@@ -1,9 +1,12 @@
use std::sync::Arc;
use axum::{extract::State, response::Response};
use axum::extract::State;
use axum::response::Response;
use http::StatusCode;
use crate::{catalog::get_dbs_and_roles, compute::ComputeNode, http::JsonResponse};
use crate::catalog::get_dbs_and_roles;
use crate::compute::ComputeNode;
use crate::http::JsonResponse;
/// Get the databases and roles from the compute.
pub(in crate::http) async fn get_catalog_objects(

View File

@@ -1,19 +1,13 @@
use std::sync::Arc;
use axum::{
extract::State,
response::{IntoResponse, Response},
};
use axum::extract::State;
use axum::response::{IntoResponse, Response};
use http::StatusCode;
use serde::Deserialize;
use crate::{
compute::ComputeNode,
http::{
extract::{Path, Query},
JsonResponse,
},
};
use crate::compute::ComputeNode;
use crate::http::JsonResponse;
use crate::http::extract::{Path, Query};
#[derive(Debug, Clone, Deserialize)]
pub(in crate::http) struct ExtensionServerParams {

View File

@@ -1,16 +1,14 @@
use std::sync::Arc;
use axum::{extract::State, response::Response};
use compute_api::{
requests::ExtensionInstallRequest,
responses::{ComputeStatus, ExtensionInstallResponse},
};
use axum::extract::State;
use axum::response::Response;
use compute_api::requests::ExtensionInstallRequest;
use compute_api::responses::{ComputeStatus, ExtensionInstallResponse};
use http::StatusCode;
use crate::{
compute::ComputeNode,
http::{extract::Json, JsonResponse},
};
use crate::compute::ComputeNode;
use crate::http::JsonResponse;
use crate::http::extract::Json;
/// Install a extension.
pub(in crate::http) async fn install_extension(

View File

@@ -17,7 +17,8 @@ pub struct FailpointConfig {
pub actions: String,
}
use crate::http::{extract::Json, JsonResponse};
use crate::http::JsonResponse;
use crate::http::extract::Json;
/// Configure failpoints for testing purposes.
pub(in crate::http) async fn configure_failpoints(

View File

@@ -1,16 +1,14 @@
use std::sync::Arc;
use axum::{extract::State, response::Response};
use compute_api::{
requests::SetRoleGrantsRequest,
responses::{ComputeStatus, SetRoleGrantsResponse},
};
use axum::extract::State;
use axum::response::Response;
use compute_api::requests::SetRoleGrantsRequest;
use compute_api::responses::{ComputeStatus, SetRoleGrantsResponse};
use http::StatusCode;
use crate::{
compute::ComputeNode,
http::{extract::Json, JsonResponse},
};
use crate::compute::ComputeNode;
use crate::http::JsonResponse;
use crate::http::extract::Json;
/// Add grants for a role.
pub(in crate::http) async fn add_grant(

View File

@@ -1,10 +1,12 @@
use std::sync::Arc;
use axum::{extract::State, response::Response};
use axum::extract::State;
use axum::response::Response;
use compute_api::responses::ComputeStatus;
use http::StatusCode;
use crate::{compute::ComputeNode, http::JsonResponse};
use crate::compute::ComputeNode;
use crate::http::JsonResponse;
/// Collect current Postgres usage insights.
pub(in crate::http) async fn get_insights(State(compute): State<Arc<ComputeNode>>) -> Response {

View File

@@ -1,10 +1,12 @@
use axum::{body::Body, response::Response};
use http::header::CONTENT_TYPE;
use axum::body::Body;
use axum::response::Response;
use http::StatusCode;
use http::header::CONTENT_TYPE;
use metrics::proto::MetricFamily;
use metrics::{Encoder, TextEncoder};
use crate::{http::JsonResponse, metrics::collect};
use crate::http::JsonResponse;
use crate::metrics::collect;
/// Expose Prometheus metrics.
pub(in crate::http) async fn get_metrics() -> Response {

View File

@@ -1,9 +1,11 @@
use std::sync::Arc;
use axum::{extract::State, response::Response};
use axum::extract::State;
use axum::response::Response;
use http::StatusCode;
use crate::{compute::ComputeNode, http::JsonResponse};
use crate::compute::ComputeNode;
use crate::http::JsonResponse;
/// Get startup metrics.
pub(in crate::http) async fn get_metrics(State(compute): State<Arc<ComputeNode>>) -> Response {

View File

@@ -1,9 +1,13 @@
use std::{ops::Deref, sync::Arc};
use std::ops::Deref;
use std::sync::Arc;
use axum::{extract::State, http::StatusCode, response::Response};
use axum::extract::State;
use axum::http::StatusCode;
use axum::response::Response;
use compute_api::responses::ComputeStatusResponse;
use crate::{compute::ComputeNode, http::JsonResponse};
use crate::compute::ComputeNode;
use crate::http::JsonResponse;
/// Retrieve the state of the comute.
pub(in crate::http) async fn get_status(State(compute): State<Arc<ComputeNode>>) -> Response {

View File

@@ -1,18 +1,14 @@
use std::sync::Arc;
use axum::{
extract::State,
response::{IntoResponse, Response},
};
use axum::extract::State;
use axum::response::{IntoResponse, Response};
use compute_api::responses::ComputeStatus;
use http::StatusCode;
use tokio::task;
use tracing::info;
use crate::{
compute::{forward_termination_signal, ComputeNode},
http::JsonResponse,
};
use crate::compute::{ComputeNode, forward_termination_signal};
use crate::http::JsonResponse;
/// Terminate the compute.
pub(in crate::http) async fn terminate(State(compute): State<Arc<ComputeNode>>) -> Response {

View File

@@ -1,23 +1,20 @@
use std::{
fmt::Display,
net::{IpAddr, Ipv6Addr, SocketAddr},
sync::Arc,
time::Duration,
};
use std::fmt::Display;
use std::net::{IpAddr, Ipv6Addr, SocketAddr};
use std::sync::Arc;
use std::time::Duration;
use anyhow::Result;
use axum::{
extract::Request,
middleware::{self, Next},
response::{IntoResponse, Response},
routing::{get, post},
Router,
};
use axum::Router;
use axum::extract::Request;
use axum::middleware::{self, Next};
use axum::response::{IntoResponse, Response};
use axum::routing::{get, post};
use http::StatusCode;
use tokio::net::TcpListener;
use tower::ServiceBuilder;
use tower_http::{request_id::PropagateRequestIdLayer, trace::TraceLayer};
use tracing::{debug, error, info, Span};
use tower_http::request_id::PropagateRequestIdLayer;
use tower_http::trace::TraceLayer;
use tracing::{Span, debug, error, info};
use uuid::Uuid;
use super::routes::{

View File

@@ -1,7 +1,7 @@
use compute_api::responses::{InstalledExtension, InstalledExtensions};
use std::collections::HashMap;
use anyhow::Result;
use compute_api::responses::{InstalledExtension, InstalledExtensions};
use postgres::{Client, NoTls};
use crate::metrics::INSTALLED_EXTENSIONS;

View File

@@ -1,17 +1,15 @@
use anyhow::bail;
use anyhow::Result;
use postgres::{NoTls, SimpleQueryMessage};
use std::time::SystemTime;
use std::{str::FromStr, sync::Arc, thread, time::Duration};
use utils::id::TenantId;
use utils::id::TimelineId;
use std::str::FromStr;
use std::sync::Arc;
use std::thread;
use std::time::{Duration, SystemTime};
use anyhow::{Result, bail};
use compute_api::spec::ComputeMode;
use postgres::{NoTls, SimpleQueryMessage};
use tracing::{info, warn};
use utils::{
lsn::Lsn,
shard::{ShardCount, ShardNumber, TenantShardId},
};
use utils::id::{TenantId, TimelineId};
use utils::lsn::Lsn;
use utils::shard::{ShardCount, ShardNumber, TenantShardId};
use crate::compute::ComputeNode;

View File

@@ -1,6 +1,6 @@
use metrics::core::Collector;
use metrics::proto::MetricFamily;
use metrics::{register_int_counter_vec, register_uint_gauge_vec, IntCounterVec, UIntGaugeVec};
use metrics::{IntCounterVec, UIntGaugeVec, register_int_counter_vec, register_uint_gauge_vec};
use once_cell::sync::Lazy;
pub(crate) static INSTALLED_EXTENSIONS: Lazy<UIntGaugeVec> = Lazy::new(|| {

View File

@@ -1,13 +1,14 @@
use std::sync::Arc;
use std::{thread, time::Duration};
use std::thread;
use std::time::Duration;
use chrono::{DateTime, Utc};
use compute_api::responses::ComputeStatus;
use compute_api::spec::ComputeFeature;
use postgres::{Client, NoTls};
use tracing::{debug, error, info, warn};
use crate::compute::ComputeNode;
use compute_api::responses::ComputeStatus;
use compute_api::spec::ComputeFeature;
const MONITOR_CHECK_INTERVAL: Duration = Duration::from_millis(500);

View File

@@ -9,7 +9,8 @@ use std::process::Child;
use std::str::FromStr;
use std::time::{Duration, Instant};
use anyhow::{bail, Result};
use anyhow::{Result, bail};
use compute_api::spec::{Database, GenericOption, GenericOptions, PgIdent, Role};
use futures::StreamExt;
use ini::Ini;
use notify::{RecursiveMode, Watcher};
@@ -21,8 +22,6 @@ use tokio_postgres;
use tokio_postgres::NoTls;
use tracing::{debug, error, info, instrument};
use compute_api::spec::{Database, GenericOption, GenericOptions, PgIdent, Role};
const POSTGRES_WAIT_TIMEOUT: Duration = Duration::from_millis(60 * 1000); // milliseconds
/// Escape a string for including it in a SQL literal.

View File

@@ -1,20 +1,20 @@
use anyhow::{anyhow, bail, Result};
use reqwest::StatusCode;
use std::fs::File;
use std::path::Path;
use tokio_postgres::Client;
use tracing::{error, info, instrument, warn};
use crate::config;
use crate::metrics::{CPlaneRequestRPC, CPLANE_REQUESTS_TOTAL, UNKNOWN_HTTP_STATUS};
use crate::migration::MigrationRunner;
use crate::params::PG_HBA_ALL_MD5;
use crate::pg_helpers::*;
use anyhow::{Result, anyhow, bail};
use compute_api::responses::{
ComputeCtlConfig, ControlPlaneComputeStatus, ControlPlaneSpecResponse,
};
use compute_api::spec::ComputeSpec;
use reqwest::StatusCode;
use tokio_postgres::Client;
use tracing::{error, info, instrument, warn};
use crate::config;
use crate::metrics::{CPLANE_REQUESTS_TOTAL, CPlaneRequestRPC, UNKNOWN_HTTP_STATUS};
use crate::migration::MigrationRunner;
use crate::params::PG_HBA_ALL_MD5;
use crate::pg_helpers::*;
// Do control plane request and return response if any. In case of error it
// returns a bool flag indicating whether it makes sense to retry the request
@@ -141,7 +141,6 @@ pub fn get_spec_from_control_plane(
/// Check `pg_hba.conf` and update if needed to allow external connections.
pub fn update_pg_hba(pgdata_path: &Path) -> Result<()> {
// XXX: consider making it a part of spec.json
info!("checking pg_hba.conf");
let pghba_path = pgdata_path.join("pg_hba.conf");
if config::line_in_file(&pghba_path, PG_HBA_ALL_MD5)? {
@@ -156,12 +155,11 @@ pub fn update_pg_hba(pgdata_path: &Path) -> Result<()> {
/// Create a standby.signal file
pub fn add_standby_signal(pgdata_path: &Path) -> Result<()> {
// XXX: consider making it a part of spec.json
info!("adding standby.signal");
let signalfile = pgdata_path.join("standby.signal");
if !signalfile.exists() {
info!("created standby.signal");
File::create(signalfile)?;
info!("created standby.signal");
} else {
info!("reused pre-existing standby.signal");
}
@@ -170,7 +168,6 @@ pub fn add_standby_signal(pgdata_path: &Path) -> Result<()> {
#[instrument(skip_all)]
pub async fn handle_neon_extension_upgrade(client: &mut Client) -> Result<()> {
info!("handle neon extension upgrade");
let query = "ALTER EXTENSION neon UPDATE";
info!("update neon extension version with query: {}", query);
client.simple_query(query).await?;

View File

@@ -1,18 +1,18 @@
use std::collections::{HashMap, HashSet};
use std::fmt::{Debug, Formatter};
use std::future::Future;
use std::iter::empty;
use std::iter::once;
use std::iter::{empty, once};
use std::sync::Arc;
use crate::compute::construct_superuser_query;
use crate::pg_helpers::{escape_literal, DatabaseExt, Escaping, GenericOptionsSearch, RoleExt};
use anyhow::{bail, Result};
use anyhow::Result;
use compute_api::spec::{ComputeFeature, ComputeSpec, Database, PgIdent, Role};
use futures::future::join_all;
use tokio::sync::RwLock;
use tokio_postgres::Client;
use tracing::{debug, info_span, Instrument};
use tracing::{Instrument, debug, info_span, warn};
use crate::compute::construct_superuser_query;
use crate::pg_helpers::{DatabaseExt, Escaping, GenericOptionsSearch, RoleExt, escape_literal};
#[derive(Clone)]
pub enum DB {
@@ -47,6 +47,11 @@ pub enum PerDatabasePhase {
DeleteDBRoleReferences,
ChangeSchemaPerms,
HandleAnonExtension,
/// This is a shared phase, used for both i) dropping dangling LR subscriptions
/// before dropping the DB, and ii) dropping all subscriptions after creating
/// a fresh branch.
/// N.B. we will skip all DBs that are not present in Postgres, invalid, or
/// have `datallowconn = false` (`restrict_conn`).
DropLogicalSubscriptions,
}
@@ -168,7 +173,7 @@ where
///
/// In the future we may generate a single stream of changes and then
/// sort/merge/batch execution, but for now this is a nice way to improve
/// batching behaviour of the commands.
/// batching behavior of the commands.
async fn get_operations<'a>(
spec: &'a ComputeSpec,
ctx: &'a RwLock<MutableApplyContext>,
@@ -451,6 +456,41 @@ async fn get_operations<'a>(
)),
}))),
ApplySpecPhase::RunInEachDatabase { db, subphase } => {
// Do some checks that user DB exists and we can access it.
//
// During the phases like DropLogicalSubscriptions, DeleteDBRoleReferences,
// which happen before dropping the DB, the current run could be a retry,
// so it's a valid case when DB is absent already. The case of
// `pg_database.datallowconn = false`/`restrict_conn` is a bit tricky, as
// in theory user can have some dangling objects there, so we will fail at
// the actual drop later. Yet, to fix that in the current code we would need
// to ALTER DATABASE, and then check back, but that even more invasive, so
// that's not what we really want to do here.
//
// For ChangeSchemaPerms, skipping DBs we cannot access is totally fine.
if let DB::UserDB(db) = db {
let databases = &ctx.read().await.dbs;
let edb = match databases.get(&db.name) {
Some(edb) => edb,
None => {
warn!(
"skipping RunInEachDatabase phase {:?}, database {} doesn't exist in PostgreSQL",
subphase, db.name
);
return Ok(Box::new(empty()));
}
};
if edb.restrict_conn || edb.invalid {
warn!(
"skipping RunInEachDatabase phase {:?}, database {} is (restrict_conn={}, invalid={})",
subphase, db.name, edb.restrict_conn, edb.invalid
);
return Ok(Box::new(empty()));
}
}
match subphase {
PerDatabasePhase::DropLogicalSubscriptions => {
match &db {
@@ -530,25 +570,12 @@ async fn get_operations<'a>(
Ok(Box::new(operations))
}
PerDatabasePhase::ChangeSchemaPerms => {
let ctx = ctx.read().await;
let databases = &ctx.dbs;
let db = match &db {
// ignore schema permissions on the system database
DB::SystemDB => return Ok(Box::new(empty())),
DB::UserDB(db) => db,
};
if databases.get(&db.name).is_none() {
bail!("database {} doesn't exist in PostgreSQL", db.name);
}
let edb = databases.get(&db.name).unwrap();
if edb.restrict_conn || edb.invalid {
return Ok(Box::new(empty()));
}
let operations = vec![
Operation {
query: format!(
@@ -566,6 +593,7 @@ async fn get_operations<'a>(
Ok(Box::new(operations))
}
// TODO: remove this completely https://github.com/neondatabase/cloud/issues/22663
PerDatabasePhase::HandleAnonExtension => {
// Only install Anon into user databases
let db = match &db {

View File

@@ -1,10 +1,11 @@
use std::path::Path;
use anyhow::{anyhow, Context};
use tracing::warn;
use anyhow::{Context, anyhow};
use tracing::{instrument, warn};
pub const RESIZE_SWAP_BIN: &str = "/neonvm/bin/resize-swap";
#[instrument]
pub fn resize_swap(size_bytes: u64) -> anyhow::Result<()> {
// run `/neonvm/bin/resize-swap --once {size_bytes}`
//

View File

@@ -1,7 +1,7 @@
#[cfg(test)]
mod config_tests {
use std::fs::{remove_file, File};
use std::fs::{File, remove_file};
use std::io::{Read, Write};
use std::path::Path;

View File

@@ -887,20 +887,6 @@ fn print_timeline(
Ok(())
}
/// Returns a map of timeline IDs to timeline_id@lsn strings.
/// Connects to the pageserver to query this information.
async fn get_timeline_infos(
env: &local_env::LocalEnv,
tenant_shard_id: &TenantShardId,
) -> Result<HashMap<TimelineId, TimelineInfo>> {
Ok(get_default_pageserver(env)
.timeline_list(tenant_shard_id)
.await?
.into_iter()
.map(|timeline_info| (timeline_info.timeline_id, timeline_info))
.collect())
}
/// Helper function to get tenant id from an optional --tenant_id option or from the config file
fn get_tenant_id(
tenant_id_arg: Option<TenantId>,
@@ -1251,12 +1237,6 @@ async fn handle_endpoint(subcmd: &EndpointCmd, env: &local_env::LocalEnv) -> Res
// TODO(sharding): this command shouldn't have to specify a shard ID: we should ask the storage controller
// where shard 0 is attached, and query there.
let tenant_shard_id = get_tenant_shard_id(args.tenant_shard_id, env)?;
let timeline_infos = get_timeline_infos(env, &tenant_shard_id)
.await
.unwrap_or_else(|e| {
eprintln!("Failed to load timeline info: {}", e);
HashMap::new()
});
let timeline_name_mappings = env.timeline_name_mappings();
@@ -1285,12 +1265,9 @@ async fn handle_endpoint(subcmd: &EndpointCmd, env: &local_env::LocalEnv) -> Res
lsn.to_string()
}
_ => {
// -> primary endpoint or hot replica
// Use the LSN at the end of the timeline.
timeline_infos
.get(&endpoint.timeline_id)
.map(|bi| bi.last_record_lsn.to_string())
.unwrap_or_else(|| "?".to_string())
// As the LSN here refers to the one that the compute is started with,
// we display nothing as it is a primary/hot standby compute.
"---".to_string()
}
};

View File

@@ -335,13 +335,21 @@ impl PageServerNode {
.map(|x| x.parse::<u64>())
.transpose()
.context("Failed to parse 'checkpoint_distance' as an integer")?,
checkpoint_timeout: settings.remove("checkpoint_timeout").map(|x| x.to_string()),
checkpoint_timeout: settings
.remove("checkpoint_timeout")
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'checkpoint_timeout' as duration")?,
compaction_target_size: settings
.remove("compaction_target_size")
.map(|x| x.parse::<u64>())
.transpose()
.context("Failed to parse 'compaction_target_size' as an integer")?,
compaction_period: settings.remove("compaction_period").map(|x| x.to_string()),
compaction_period: settings
.remove("compaction_period")
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'compaction_period' as duration")?,
compaction_threshold: settings
.remove("compaction_threshold")
.map(|x| x.parse::<usize>())
@@ -387,7 +395,10 @@ impl PageServerNode {
.map(|x| x.parse::<u64>())
.transpose()
.context("Failed to parse 'gc_horizon' as an integer")?,
gc_period: settings.remove("gc_period").map(|x| x.to_string()),
gc_period: settings.remove("gc_period")
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'gc_period' as duration")?,
image_creation_threshold: settings
.remove("image_creation_threshold")
.map(|x| x.parse::<usize>())
@@ -403,13 +414,20 @@ impl PageServerNode {
.map(|x| x.parse::<usize>())
.transpose()
.context("Failed to parse 'image_creation_preempt_threshold' as integer")?,
pitr_interval: settings.remove("pitr_interval").map(|x| x.to_string()),
pitr_interval: settings.remove("pitr_interval")
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'pitr_interval' as duration")?,
walreceiver_connect_timeout: settings
.remove("walreceiver_connect_timeout")
.map(|x| x.to_string()),
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'walreceiver_connect_timeout' as duration")?,
lagging_wal_timeout: settings
.remove("lagging_wal_timeout")
.map(|x| x.to_string()),
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'lagging_wal_timeout' as duration")?,
max_lsn_wal_lag: settings
.remove("max_lsn_wal_lag")
.map(|x| x.parse::<NonZeroU64>())
@@ -427,8 +445,14 @@ impl PageServerNode {
.context("Failed to parse 'min_resident_size_override' as integer")?,
evictions_low_residence_duration_metric_threshold: settings
.remove("evictions_low_residence_duration_metric_threshold")
.map(|x| x.to_string()),
heatmap_period: settings.remove("heatmap_period").map(|x| x.to_string()),
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'evictions_low_residence_duration_metric_threshold' as duration")?,
heatmap_period: settings
.remove("heatmap_period")
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'heatmap_period' as duration")?,
lazy_slru_download: settings
.remove("lazy_slru_download")
.map(|x| x.parse::<bool>())
@@ -439,10 +463,15 @@ impl PageServerNode {
.map(serde_json::from_str)
.transpose()
.context("parse `timeline_get_throttle` from json")?,
lsn_lease_length: settings.remove("lsn_lease_length").map(|x| x.to_string()),
lsn_lease_length: settings.remove("lsn_lease_length")
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'lsn_lease_length' as duration")?,
lsn_lease_length_for_ts: settings
.remove("lsn_lease_length_for_ts")
.map(|x| x.to_string()),
.map(humantime::parse_duration)
.transpose()
.context("Failed to parse 'lsn_lease_length_for_ts' as duration")?,
timeline_offloading: settings
.remove("timeline_offloading")
.map(|x| x.parse::<bool>())

View File

@@ -959,7 +959,7 @@ async fn main() -> anyhow::Result<()> {
threshold: threshold.into(),
},
)),
heatmap_period: Some("300s".to_string()),
heatmap_period: Some(Duration::from_secs(300)),
..Default::default()
},
})

View File

@@ -51,8 +51,6 @@ for pg_version in ${TEST_VERSION_ONLY-14 15 16 17}; do
done
if [ $pg_version -ge 16 ]; then
docker cp ext-src $TEST_CONTAINER_NAME:/
docker exec $TEST_CONTAINER_NAME bash -c "apt update && apt install -y libtap-parser-sourcehandler-pgtap-perl"
# This is required for the pg_hint_plan test, to prevent flaky log message causing the test to fail
# It cannot be moved to Dockerfile now because the database directory is created after the start of the container
echo Adding dummy config

View File

@@ -57,7 +57,6 @@ new_vers=$(docker compose exec neon-test-extensions psql -Aqt -d contrib_regress
docker compose --profile test-extensions down
TAG=${OLDTAG} docker compose --profile test-extensions up --quiet-pull --build -d --force-recreate
wait_for_ready
docker compose cp ext-src neon-test-extensions:/
docker compose exec neon-test-extensions psql -c "DROP DATABASE IF EXISTS contrib_regression"
docker compose exec neon-test-extensions psql -c "CREATE DATABASE contrib_regression"
docker compose exec neon-test-extensions psql -c "CREATE DATABASE pgtap_regression"

View File

@@ -1,7 +1,7 @@
[package]
name = "compute_api"
version = "0.1.0"
edition.workspace = true
edition = "2024"
license.workspace = true
[dependencies]

View File

@@ -1,11 +1,10 @@
//! Structs representing the JSON formats used in the compute_ctl's HTTP API.
use crate::{
privilege::Privilege,
responses::ComputeCtlConfig,
spec::{ComputeSpec, ExtVersion, PgIdent},
};
use serde::{Deserialize, Serialize};
use crate::privilege::Privilege;
use crate::responses::ComputeCtlConfig;
use crate::spec::{ComputeSpec, ExtVersion, PgIdent};
/// Request of the /configure API
///
/// We now pass only `spec` in the configuration request, but later we can

View File

@@ -6,10 +6,8 @@ use chrono::{DateTime, Utc};
use jsonwebtoken::jwk::JwkSet;
use serde::{Deserialize, Serialize, Serializer};
use crate::{
privilege::Privilege,
spec::{ComputeSpec, Database, ExtVersion, PgIdent, Role},
};
use crate::privilege::Privilege;
use crate::spec::{ComputeSpec, Database, ExtVersion, PgIdent, Role};
#[derive(Serialize, Debug, Deserialize)]
pub struct GenericAPIError {

View File

@@ -5,13 +5,12 @@
//! and connect it to the storage nodes.
use std::collections::HashMap;
use regex::Regex;
use remote_storage::RemotePath;
use serde::{Deserialize, Serialize};
use utils::id::{TenantId, TimelineId};
use utils::lsn::Lsn;
use regex::Regex;
use remote_storage::RemotePath;
/// String type alias representing Postgres identifier and
/// intended to be used for DB / role names.
pub type PgIdent = String;
@@ -339,9 +338,10 @@ pub struct JwksSettings {
#[cfg(test)]
mod tests {
use super::*;
use std::fs::File;
use super::*;
#[test]
fn allow_installing_remote_extensions() {
let rspec: RemoteExtSpec = serde_json::from_value(serde_json::json!({

View File

@@ -1,7 +1,7 @@
[package]
name = "pageserver_api"
version = "0.1.0"
edition.workspace = true
edition = "2024"
license.workspace = true
[features]

View File

@@ -9,19 +9,18 @@ pub const DEFAULT_PG_LISTEN_ADDR: &str = formatcp!("127.0.0.1:{DEFAULT_PG_LISTEN
pub const DEFAULT_HTTP_LISTEN_PORT: u16 = 9898;
pub const DEFAULT_HTTP_LISTEN_ADDR: &str = formatcp!("127.0.0.1:{DEFAULT_HTTP_LISTEN_PORT}");
use std::collections::HashMap;
use std::num::{NonZeroU64, NonZeroUsize};
use std::str::FromStr;
use std::time::Duration;
use postgres_backend::AuthType;
use remote_storage::RemoteStorageConfig;
use serde_with::serde_as;
use std::{
collections::HashMap,
num::{NonZeroU64, NonZeroUsize},
str::FromStr,
time::Duration,
};
use utils::{logging::LogFormat, postgres_client::PostgresClientProtocol};
use utils::logging::LogFormat;
use utils::postgres_client::PostgresClientProtocol;
use crate::models::ImageCompressionAlgorithm;
use crate::models::LsnLease;
use crate::models::{ImageCompressionAlgorithm, LsnLease};
// Certain metadata (e.g. externally-addressable name, AZ) is delivered
// as a separate structure. This information is not neeed by the pageserver
@@ -367,10 +366,10 @@ pub struct TenantConfigToml {
}
pub mod defaults {
use crate::models::ImageCompressionAlgorithm;
pub use storage_broker::DEFAULT_ENDPOINT as BROKER_DEFAULT_ENDPOINT;
use crate::models::ImageCompressionAlgorithm;
pub const DEFAULT_WAIT_LSN_TIMEOUT: &str = "300 s";
pub const DEFAULT_WAL_REDO_TIMEOUT: &str = "60 s";
@@ -584,7 +583,7 @@ pub mod tenant_conf_defaults {
// image layers should be created.
pub const DEFAULT_IMAGE_LAYER_CREATION_CHECK_THRESHOLD: u8 = 2;
pub const DEFAULT_GC_COMPACTION_ENABLED: bool = false;
pub const DEFAULT_GC_COMPACTION_INITIAL_THRESHOLD_KB: u64 = 10240000;
pub const DEFAULT_GC_COMPACTION_INITIAL_THRESHOLD_KB: u64 = 5 * 1024 * 1024; // 5GB
pub const DEFAULT_GC_COMPACTION_RATIO_PERCENT: u64 = 100;
}

View File

@@ -9,11 +9,8 @@ use std::time::{Duration, Instant};
use serde::{Deserialize, Serialize};
use utils::id::{NodeId, TenantId};
use crate::models::PageserverUtilization;
use crate::{
models::{ShardParameters, TenantConfig},
shard::{ShardStripeSize, TenantShardId},
};
use crate::models::{PageserverUtilization, ShardParameters, TenantConfig};
use crate::shard::{ShardStripeSize, TenantShardId};
#[derive(Serialize, Deserialize, Debug)]
#[serde(deny_unknown_fields)]
@@ -354,7 +351,7 @@ impl FromStr for SkSchedulingPolicy {
_ => {
return Err(anyhow::anyhow!(
"Unknown scheduling policy '{s}', try active,pause,decomissioned"
))
));
}
})
}
@@ -457,9 +454,10 @@ pub struct SafekeeperSchedulingPolicyRequest {
#[cfg(test)]
mod test {
use super::*;
use serde_json;
use super::*;
/// Check stability of PlacementPolicy's serialization
#[test]
fn placement_policy_encoding() -> anyhow::Result<()> {

View File

@@ -1,11 +1,12 @@
use anyhow::{bail, Result};
use byteorder::{ByteOrder, BE};
use std::fmt;
use std::ops::Range;
use anyhow::{Result, bail};
use byteorder::{BE, ByteOrder};
use bytes::Bytes;
use postgres_ffi::relfile_utils::{FSM_FORKNUM, VISIBILITYMAP_FORKNUM};
use postgres_ffi::Oid;
use postgres_ffi::RepOriginId;
use postgres_ffi::{Oid, RepOriginId};
use serde::{Deserialize, Serialize};
use std::{fmt, ops::Range};
use utils::const_assert;
use crate::reltag::{BlockNumber, RelTag, SlruKind};
@@ -954,25 +955,22 @@ impl std::str::FromStr for Key {
mod tests {
use std::str::FromStr;
use crate::key::is_metadata_key_slice;
use crate::key::Key;
use rand::Rng;
use rand::SeedableRng;
use rand::{Rng, SeedableRng};
use super::AUX_KEY_PREFIX;
use crate::key::{Key, is_metadata_key_slice};
#[test]
fn display_fromstr_bijection() {
let mut rng = rand::rngs::StdRng::seed_from_u64(42);
let key = Key {
field1: rng.gen(),
field2: rng.gen(),
field3: rng.gen(),
field4: rng.gen(),
field5: rng.gen(),
field6: rng.gen(),
field1: rng.r#gen(),
field2: rng.r#gen(),
field3: rng.r#gen(),
field4: rng.r#gen(),
field5: rng.r#gen(),
field6: rng.r#gen(),
};
assert_eq!(key, Key::from_str(&format!("{key}")).unwrap());

View File

@@ -1,11 +1,10 @@
use postgres_ffi::BLCKSZ;
use std::ops::Range;
use crate::{
key::Key,
shard::{ShardCount, ShardIdentity},
};
use itertools::Itertools;
use postgres_ffi::BLCKSZ;
use crate::key::Key;
use crate::shard::{ShardCount, ShardIdentity};
///
/// Represents a set of Keys, in a compact form.
@@ -609,15 +608,13 @@ pub fn singleton_range(key: Key) -> Range<Key> {
#[cfg(test)]
mod tests {
use std::fmt::Write;
use rand::{RngCore, SeedableRng};
use crate::{
models::ShardParameters,
shard::{ShardCount, ShardNumber},
};
use super::*;
use std::fmt::Write;
use crate::models::ShardParameters;
use crate::shard::{ShardCount, ShardNumber};
// Helper function to create a key range.
//

View File

@@ -2,38 +2,30 @@ pub mod detach_ancestor;
pub mod partitioning;
pub mod utilization;
#[cfg(feature = "testing")]
use camino::Utf8PathBuf;
pub use utilization::PageserverUtilization;
use core::ops::Range;
use std::{
collections::HashMap,
fmt::Display,
io::{BufRead, Read},
num::{NonZeroU32, NonZeroU64, NonZeroUsize},
str::FromStr,
time::{Duration, SystemTime},
};
use std::collections::HashMap;
use std::fmt::Display;
use std::io::{BufRead, Read};
use std::num::{NonZeroU32, NonZeroU64, NonZeroUsize};
use std::str::FromStr;
use std::time::{Duration, SystemTime};
use byteorder::{BigEndian, ReadBytesExt};
use bytes::{Buf, BufMut, Bytes, BytesMut};
#[cfg(feature = "testing")]
use camino::Utf8PathBuf;
use postgres_ffi::BLCKSZ;
use serde::{Deserialize, Deserializer, Serialize, Serializer};
use serde_with::serde_as;
use utils::{
completion,
id::{NodeId, TenantId, TimelineId},
lsn::Lsn,
postgres_client::PostgresClientProtocol,
serde_system_time,
};
pub use utilization::PageserverUtilization;
use utils::id::{NodeId, TenantId, TimelineId};
use utils::lsn::Lsn;
use utils::postgres_client::PostgresClientProtocol;
use utils::{completion, serde_system_time};
use crate::{
key::{CompactKey, Key},
reltag::RelTag,
shard::{ShardCount, ShardStripeSize, TenantShardId},
};
use bytes::{Buf, BufMut, Bytes, BytesMut};
use crate::key::{CompactKey, Key};
use crate::reltag::RelTag;
use crate::shard::{ShardCount, ShardStripeSize, TenantShardId};
/// The state of a tenant in this pageserver.
///
@@ -332,7 +324,8 @@ pub struct ImportPgdataIdempotencyKey(pub String);
impl ImportPgdataIdempotencyKey {
pub fn random() -> Self {
use rand::{distributions::Alphanumeric, Rng};
use rand::Rng;
use rand::distributions::Alphanumeric;
Self(
rand::thread_rng()
.sample_iter(&Alphanumeric)
@@ -526,9 +519,13 @@ pub struct TenantConfigPatch {
#[derive(Serialize, Deserialize, Debug, Default, Clone, Eq, PartialEq)]
pub struct TenantConfig {
pub checkpoint_distance: Option<u64>,
pub checkpoint_timeout: Option<String>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub checkpoint_timeout: Option<Duration>,
pub compaction_target_size: Option<u64>,
pub compaction_period: Option<String>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub compaction_period: Option<Duration>,
pub compaction_threshold: Option<usize>,
pub compaction_upper_limit: Option<usize>,
// defer parsing compaction_algorithm, like eviction_policy
@@ -539,22 +536,38 @@ pub struct TenantConfig {
pub l0_flush_stall_threshold: Option<usize>,
pub l0_flush_wait_upload: Option<bool>,
pub gc_horizon: Option<u64>,
pub gc_period: Option<String>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub gc_period: Option<Duration>,
pub image_creation_threshold: Option<usize>,
pub pitr_interval: Option<String>,
pub walreceiver_connect_timeout: Option<String>,
pub lagging_wal_timeout: Option<String>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub pitr_interval: Option<Duration>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub walreceiver_connect_timeout: Option<Duration>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub lagging_wal_timeout: Option<Duration>,
pub max_lsn_wal_lag: Option<NonZeroU64>,
pub eviction_policy: Option<EvictionPolicy>,
pub min_resident_size_override: Option<u64>,
pub evictions_low_residence_duration_metric_threshold: Option<String>,
pub heatmap_period: Option<String>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub evictions_low_residence_duration_metric_threshold: Option<Duration>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub heatmap_period: Option<Duration>,
pub lazy_slru_download: Option<bool>,
pub timeline_get_throttle: Option<ThrottleConfig>,
pub image_layer_creation_check_threshold: Option<u8>,
pub image_creation_preempt_threshold: Option<usize>,
pub lsn_lease_length: Option<String>,
pub lsn_lease_length_for_ts: Option<String>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub lsn_lease_length: Option<Duration>,
#[serde(default)]
#[serde(with = "humantime_serde")]
pub lsn_lease_length_for_ts: Option<Duration>,
pub timeline_offloading: Option<bool>,
pub wal_receiver_protocol_override: Option<PostgresClientProtocol>,
pub rel_size_v2_enabled: Option<bool>,
@@ -564,7 +577,10 @@ pub struct TenantConfig {
}
impl TenantConfig {
pub fn apply_patch(self, patch: TenantConfigPatch) -> TenantConfig {
pub fn apply_patch(
self,
patch: TenantConfigPatch,
) -> Result<TenantConfig, humantime::DurationError> {
let Self {
mut checkpoint_distance,
mut checkpoint_timeout,
@@ -604,11 +620,17 @@ impl TenantConfig {
} = self;
patch.checkpoint_distance.apply(&mut checkpoint_distance);
patch.checkpoint_timeout.apply(&mut checkpoint_timeout);
patch
.checkpoint_timeout
.map(|v| humantime::parse_duration(&v))?
.apply(&mut checkpoint_timeout);
patch
.compaction_target_size
.apply(&mut compaction_target_size);
patch.compaction_period.apply(&mut compaction_period);
patch
.compaction_period
.map(|v| humantime::parse_duration(&v))?
.apply(&mut compaction_period);
patch.compaction_threshold.apply(&mut compaction_threshold);
patch
.compaction_upper_limit
@@ -626,15 +648,25 @@ impl TenantConfig {
.apply(&mut l0_flush_stall_threshold);
patch.l0_flush_wait_upload.apply(&mut l0_flush_wait_upload);
patch.gc_horizon.apply(&mut gc_horizon);
patch.gc_period.apply(&mut gc_period);
patch
.gc_period
.map(|v| humantime::parse_duration(&v))?
.apply(&mut gc_period);
patch
.image_creation_threshold
.apply(&mut image_creation_threshold);
patch.pitr_interval.apply(&mut pitr_interval);
patch
.pitr_interval
.map(|v| humantime::parse_duration(&v))?
.apply(&mut pitr_interval);
patch
.walreceiver_connect_timeout
.map(|v| humantime::parse_duration(&v))?
.apply(&mut walreceiver_connect_timeout);
patch.lagging_wal_timeout.apply(&mut lagging_wal_timeout);
patch
.lagging_wal_timeout
.map(|v| humantime::parse_duration(&v))?
.apply(&mut lagging_wal_timeout);
patch.max_lsn_wal_lag.apply(&mut max_lsn_wal_lag);
patch.eviction_policy.apply(&mut eviction_policy);
patch
@@ -642,8 +674,12 @@ impl TenantConfig {
.apply(&mut min_resident_size_override);
patch
.evictions_low_residence_duration_metric_threshold
.map(|v| humantime::parse_duration(&v))?
.apply(&mut evictions_low_residence_duration_metric_threshold);
patch.heatmap_period.apply(&mut heatmap_period);
patch
.heatmap_period
.map(|v| humantime::parse_duration(&v))?
.apply(&mut heatmap_period);
patch.lazy_slru_download.apply(&mut lazy_slru_download);
patch
.timeline_get_throttle
@@ -654,9 +690,13 @@ impl TenantConfig {
patch
.image_creation_preempt_threshold
.apply(&mut image_creation_preempt_threshold);
patch.lsn_lease_length.apply(&mut lsn_lease_length);
patch
.lsn_lease_length
.map(|v| humantime::parse_duration(&v))?
.apply(&mut lsn_lease_length);
patch
.lsn_lease_length_for_ts
.map(|v| humantime::parse_duration(&v))?
.apply(&mut lsn_lease_length_for_ts);
patch.timeline_offloading.apply(&mut timeline_offloading);
patch
@@ -673,7 +713,7 @@ impl TenantConfig {
.gc_compaction_ratio_percent
.apply(&mut gc_compaction_ratio_percent);
Self {
Ok(Self {
checkpoint_distance,
checkpoint_timeout,
compaction_target_size,
@@ -709,7 +749,7 @@ impl TenantConfig {
gc_compaction_enabled,
gc_compaction_initial_threshold_kb,
gc_compaction_ratio_percent,
}
})
}
}
@@ -2241,9 +2281,10 @@ impl Default for PageTraceEvent {
#[cfg(test)]
mod tests {
use serde_json::json;
use std::str::FromStr;
use serde_json::json;
use super::*;
#[test]
@@ -2503,7 +2544,7 @@ mod tests {
..base.clone()
};
let patched = base.apply_patch(decoded.config);
let patched = base.apply_patch(decoded.config).unwrap();
assert_eq!(patched, expected);
}

View File

@@ -1,5 +1,7 @@
use std::time::SystemTime;
use utils::{serde_percent::Percent, serde_system_time};
use utils::serde_percent::Percent;
use utils::serde_system_time;
/// Pageserver current utilization and scoring for how good candidate the pageserver would be for
/// the next tenant.
@@ -131,12 +133,12 @@ impl PageserverUtilization {
/// Test helper
pub mod test_utilization {
use super::PageserverUtilization;
use std::time::SystemTime;
use utils::{
serde_percent::Percent,
serde_system_time::{self},
};
use utils::serde_percent::Percent;
use utils::serde_system_time::{self};
use super::PageserverUtilization;
// Parameters of the imaginary node used for test utilization instances
const TEST_DISK_SIZE: u64 = 1024 * 1024 * 1024 * 1024;

View File

@@ -1,7 +1,7 @@
//! This module defines the WAL record format used within the pageserver.
use bytes::Bytes;
use postgres_ffi::walrecord::{describe_postgres_wal_record, MultiXactMember};
use postgres_ffi::walrecord::{MultiXactMember, describe_postgres_wal_record};
use postgres_ffi::{MultiXactId, MultiXactOffset, TimestampTz, TransactionId};
use serde::{Deserialize, Serialize};
use utils::bin_ser::DeserializeError;

View File

@@ -1,10 +1,10 @@
use serde::{Deserialize, Serialize};
use std::cmp::Ordering;
use std::fmt;
use postgres_ffi::pg_constants::GLOBALTABLESPACE_OID;
use postgres_ffi::relfile_utils::{forkname_to_number, forknumber_to_name, MAIN_FORKNUM};
use postgres_ffi::Oid;
use postgres_ffi::pg_constants::GLOBALTABLESPACE_OID;
use postgres_ffi::relfile_utils::{MAIN_FORKNUM, forkname_to_number, forknumber_to_name};
use serde::{Deserialize, Serialize};
///
/// Relation data file segment id throughout the Postgres cluster.

View File

@@ -33,12 +33,13 @@
use std::hash::{Hash, Hasher};
use crate::{key::Key, models::ShardParameters};
#[doc(inline)]
pub use ::utils::shard::*;
use postgres_ffi::relfile_utils::INIT_FORKNUM;
use serde::{Deserialize, Serialize};
#[doc(inline)]
pub use ::utils::shard::*;
use crate::key::Key;
use crate::models::ShardParameters;
/// The ShardIdentity contains enough information to map a [`Key`] to a [`ShardNumber`],
/// and to check whether that [`ShardNumber`] is the same as the current shard.
@@ -337,7 +338,8 @@ pub fn describe(
mod tests {
use std::str::FromStr;
use utils::{id::TenantId, Hex};
use utils::Hex;
use utils::id::TenantId;
use super::*;

View File

@@ -6,9 +6,9 @@
use serde::{Deserialize, Serialize};
use utils::id::NodeId;
use crate::{
controller_api::NodeRegisterRequest, models::LocationConfigMode, shard::TenantShardId,
};
use crate::controller_api::NodeRegisterRequest;
use crate::models::LocationConfigMode;
use crate::shard::TenantShardId;
/// Upcall message sent by the pageserver to the configured `control_plane_api` on
/// startup.
@@ -30,7 +30,7 @@ fn default_mode() -> LocationConfigMode {
pub struct ReAttachResponseTenant {
pub id: TenantShardId,
/// Mandatory if LocationConfigMode is None or set to an Attached* mode
pub gen: Option<u32>,
pub r#gen: Option<u32>,
/// Default value only for backward compat: this field should be set
#[serde(default = "default_mode")]
@@ -44,7 +44,7 @@ pub struct ReAttachResponse {
#[derive(Serialize, Deserialize)]
pub struct ValidateRequestTenant {
pub id: TenantShardId,
pub gen: u32,
pub r#gen: u32,
}
#[derive(Serialize, Deserialize)]

View File

@@ -7,10 +7,11 @@
//! Note that the [`Value`] type is used for the permananent storage format, so any
//! changes to it must be backwards compatible.
use crate::record::NeonWalRecord;
use bytes::Bytes;
use serde::{Deserialize, Serialize};
use crate::record::NeonWalRecord;
#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, Eq)]
pub enum Value {
/// An Image value contains a full copy of the value
@@ -83,11 +84,11 @@ impl ValueBytes {
#[cfg(test)]
mod test {
use super::*;
use bytes::Bytes;
use utils::bin_ser::BeSer;
use super::*;
macro_rules! roundtrip {
($orig:expr, $expected:expr) => {{
let orig: Value = $orig;

View File

@@ -7,7 +7,6 @@ license.workspace = true
[dependencies]
anyhow.workspace = true
itertools.workspace = true
postgres.workspace = true
tokio-postgres.workspace = true
url.workspace = true

View File

@@ -171,10 +171,10 @@ impl PgConnectionConfig {
tokio_postgres::Client,
tokio_postgres::Connection<tokio_postgres::Socket, tokio_postgres::tls::NoTlsStream>,
),
postgres::Error,
tokio_postgres::Error,
> {
self.to_tokio_postgres_config()
.connect(postgres::NoTls)
.connect(tokio_postgres::NoTls)
.await
}
}

View File

@@ -1,7 +1,7 @@
[package]
name = "postgres-protocol2"
version = "0.1.0"
edition = "2021"
edition = "2024"
license = "MIT/Apache-2.0"
[dependencies]

View File

@@ -1,14 +1,12 @@
//! SASL-based authentication support.
use std::fmt::Write;
use std::{io, iter, mem, str};
use hmac::{Hmac, Mac};
use rand::{self, Rng};
use sha2::digest::FixedOutput;
use sha2::{Digest, Sha256};
use std::fmt::Write;
use std::io;
use std::iter;
use std::mem;
use std::str;
use tokio::task::yield_now;
const NONCE_LENGTH: usize = 24;
@@ -493,11 +491,9 @@ mod test {
let nonce = "9IZ2O01zb9IgiIZ1WJ/zgpJB";
let client_first = "n,,n=,r=9IZ2O01zb9IgiIZ1WJ/zgpJB";
let server_first =
"r=9IZ2O01zb9IgiIZ1WJ/zgpJBjx/oIRLs02gGSHcw1KEty3eY,s=fs3IXBy7U7+IvVjZ,i\
let server_first = "r=9IZ2O01zb9IgiIZ1WJ/zgpJBjx/oIRLs02gGSHcw1KEty3eY,s=fs3IXBy7U7+IvVjZ,i\
=4096";
let client_final =
"c=biws,r=9IZ2O01zb9IgiIZ1WJ/zgpJBjx/oIRLs02gGSHcw1KEty3eY,p=AmNKosjJzS3\
let client_final = "c=biws,r=9IZ2O01zb9IgiIZ1WJ/zgpJBjx/oIRLs02gGSHcw1KEty3eY,p=AmNKosjJzS3\
1NTlQYNs5BTeQjdHdk7lOflDo5re2an8=";
let server_final = "v=U+ppxD5XUKtradnv8e2MkeupiA8FU87Sg8CXzXHDAzw=";

View File

@@ -11,9 +11,10 @@
//! set to `UTF8`. It will most likely not behave properly if that is not the case.
#![warn(missing_docs, clippy::all)]
use std::io;
use byteorder::{BigEndian, ByteOrder};
use bytes::{BufMut, BytesMut};
use std::io;
pub mod authentication;
pub mod escape;

View File

@@ -1,13 +1,13 @@
#![allow(missing_docs)]
use std::io::{self, Read};
use std::ops::Range;
use std::{cmp, str};
use byteorder::{BigEndian, ByteOrder, ReadBytesExt};
use bytes::{Bytes, BytesMut};
use fallible_iterator::FallibleIterator;
use memchr::memchr;
use std::cmp;
use std::io::{self, Read};
use std::ops::Range;
use std::str;
use crate::Oid;

View File

@@ -1,13 +1,13 @@
//! Frontend message serialization.
#![allow(missing_docs)]
use std::error::Error;
use std::{io, marker};
use byteorder::{BigEndian, ByteOrder};
use bytes::{Buf, BufMut, BytesMut};
use std::error::Error;
use std::io;
use std::marker;
use crate::{write_nullable, FromUsize, IsNull, Oid};
use crate::{FromUsize, IsNull, Oid, write_nullable};
#[inline]
fn write_body<F, E>(buf: &mut BytesMut, f: F) -> Result<(), E>

View File

@@ -6,12 +6,13 @@
//! side. This is good because it ensures the cleartext password won't
//! end up in logs pg_stat displays, etc.
use crate::authentication::sasl;
use hmac::{Hmac, Mac};
use rand::RngCore;
use sha2::digest::FixedOutput;
use sha2::{Digest, Sha256};
use crate::authentication::sasl;
#[cfg(test)]
mod test;

View File

@@ -1,11 +1,12 @@
//! Conversions to and from Postgres's binary format for various types.
use byteorder::{BigEndian, ReadBytesExt};
use bytes::{BufMut, BytesMut};
use fallible_iterator::FallibleIterator;
use std::boxed::Box as StdBox;
use std::error::Error;
use std::str;
use byteorder::{BigEndian, ReadBytesExt};
use bytes::{BufMut, BytesMut};
use fallible_iterator::FallibleIterator;
use crate::Oid;
#[cfg(test)]

View File

@@ -1,7 +1,7 @@
[package]
name = "postgres-types2"
version = "0.1.0"
edition = "2021"
edition = "2024"
license = "MIT/Apache-2.0"
[dependencies]

View File

@@ -4,19 +4,18 @@
//! unless you want to define your own `ToSql` or `FromSql` definitions.
#![warn(clippy::all, missing_docs)]
use fallible_iterator::FallibleIterator;
use postgres_protocol2::types;
use std::any::type_name;
use std::error::Error;
use std::fmt;
use std::sync::Arc;
use crate::type_gen::{Inner, Other};
use bytes::BytesMut;
use fallible_iterator::FallibleIterator;
#[doc(inline)]
pub use postgres_protocol2::Oid;
use postgres_protocol2::types;
use bytes::BytesMut;
use crate::type_gen::{Inner, Other};
/// Generates a simple implementation of `ToSql::accepts` which accepts the
/// types passed to it.

View File

@@ -1,7 +1,9 @@
use crate::{FromSql, Type};
pub use bytes::BytesMut;
use std::error::Error;
pub use bytes::BytesMut;
use crate::{FromSql, Type};
pub fn read_be_i32(buf: &mut &[u8]) -> Result<i32, Box<dyn Error + Sync + Send>> {
if buf.len() < 4 {
return Err("invalid buffer size".into());

View File

@@ -1,7 +1,7 @@
[package]
name = "tokio-postgres2"
version = "0.1.0"
edition = "2021"
edition = "2024"
license = "MIT/Apache-2.0"
[dependencies]

View File

@@ -1,10 +1,11 @@
use std::io;
use tokio::net::TcpStream;
use crate::client::SocketConfig;
use crate::config::{Host, SslMode};
use crate::tls::MakeTlsConnect;
use crate::{cancel_query_raw, connect_socket, Error};
use std::io;
use crate::{Error, cancel_query_raw, connect_socket};
pub(crate) async fn cancel_query<T>(
config: Option<SocketConfig>,
@@ -22,7 +23,7 @@ where
return Err(Error::connect(io::Error::new(
io::ErrorKind::InvalidInput,
"unknown host",
)))
)));
}
};

View File

@@ -1,10 +1,11 @@
use crate::config::SslMode;
use crate::tls::TlsConnect;
use crate::{connect_tls, Error};
use bytes::BytesMut;
use postgres_protocol2::message::frontend;
use tokio::io::{AsyncRead, AsyncWrite, AsyncWriteExt};
use crate::config::SslMode;
use crate::tls::TlsConnect;
use crate::{Error, connect_tls};
pub async fn cancel_query_raw<S, T>(
stream: S,
mode: SslMode,

View File

@@ -1,12 +1,12 @@
use crate::config::SslMode;
use crate::tls::TlsConnect;
use crate::{cancel_query, client::SocketConfig, tls::MakeTlsConnect};
use crate::{cancel_query_raw, Error};
use serde::{Deserialize, Serialize};
use tokio::io::{AsyncRead, AsyncWrite};
use tokio::net::TcpStream;
use crate::client::SocketConfig;
use crate::config::SslMode;
use crate::tls::{MakeTlsConnect, TlsConnect};
use crate::{Error, cancel_query, cancel_query_raw};
/// The capability to request cancellation of in-progress queries on a
/// connection.
#[derive(Clone, Serialize, Deserialize)]

View File

@@ -1,31 +1,28 @@
use crate::codec::{BackendMessages, FrontendMessage};
use crate::config::Host;
use crate::config::SslMode;
use crate::connection::{Request, RequestMessages};
use crate::query::RowStream;
use crate::simple_query::SimpleQueryStream;
use crate::types::{Oid, ToSql, Type};
use crate::{
query, simple_query, slice_iter, CancelToken, Error, ReadyForQueryStatus, Row,
SimpleQueryMessage, Statement, Transaction, TransactionBuilder,
};
use bytes::BytesMut;
use fallible_iterator::FallibleIterator;
use futures_util::{future, ready, TryStreamExt};
use parking_lot::Mutex;
use postgres_protocol2::message::{backend::Message, frontend};
use serde::{Deserialize, Serialize};
use std::collections::HashMap;
use std::fmt;
use std::sync::Arc;
use std::task::{Context, Poll};
use std::time::Duration;
use bytes::BytesMut;
use fallible_iterator::FallibleIterator;
use futures_util::{TryStreamExt, future, ready};
use parking_lot::Mutex;
use postgres_protocol2::message::backend::Message;
use postgres_protocol2::message::frontend;
use serde::{Deserialize, Serialize};
use tokio::sync::mpsc;
use std::time::Duration;
use crate::codec::{BackendMessages, FrontendMessage};
use crate::config::{Host, SslMode};
use crate::connection::{Request, RequestMessages};
use crate::query::RowStream;
use crate::simple_query::SimpleQueryStream;
use crate::types::{Oid, ToSql, Type};
use crate::{
CancelToken, Error, ReadyForQueryStatus, Row, SimpleQueryMessage, Statement, Transaction,
TransactionBuilder, query, simple_query, slice_iter,
};
pub struct Responses {
receiver: mpsc::Receiver<BackendMessages>,

View File

@@ -1,8 +1,9 @@
use std::io;
use bytes::{Buf, Bytes, BytesMut};
use fallible_iterator::FallibleIterator;
use postgres_protocol2::message::backend;
use postgres_protocol2::message::frontend::CopyData;
use std::io;
use tokio_util::codec::{Decoder, Encoder};
pub enum FrontendMessage {

View File

@@ -1,21 +1,19 @@
//! Connection configuration.
use crate::connect::connect;
use crate::connect_raw::connect_raw;
use crate::connect_raw::RawConnection;
use crate::tls::MakeTlsConnect;
use crate::tls::TlsConnect;
use crate::{Client, Connection, Error};
use postgres_protocol2::message::frontend::StartupMessageParams;
use serde::{Deserialize, Serialize};
use std::fmt;
use std::str;
use std::time::Duration;
use tokio::io::{AsyncRead, AsyncWrite};
use std::{fmt, str};
pub use postgres_protocol2::authentication::sasl::ScramKeys;
use postgres_protocol2::message::frontend::StartupMessageParams;
use serde::{Deserialize, Serialize};
use tokio::io::{AsyncRead, AsyncWrite};
use tokio::net::TcpStream;
use crate::connect::connect;
use crate::connect_raw::{RawConnection, connect_raw};
use crate::tls::{MakeTlsConnect, TlsConnect};
use crate::{Client, Connection, Error};
/// TLS configuration.
#[derive(Debug, Copy, Clone, PartialEq, Eq, Serialize, Deserialize)]
#[non_exhaustive]

View File

@@ -1,3 +1,7 @@
use postgres_protocol2::message::backend::Message;
use tokio::net::TcpStream;
use tokio::sync::mpsc;
use crate::client::SocketConfig;
use crate::codec::BackendMessage;
use crate::config::Host;
@@ -5,9 +9,6 @@ use crate::connect_raw::connect_raw;
use crate::connect_socket::connect_socket;
use crate::tls::{MakeTlsConnect, TlsConnect};
use crate::{Client, Config, Connection, Error, RawConnection};
use postgres_protocol2::message::backend::Message;
use tokio::net::TcpStream;
use tokio::sync::mpsc;
pub async fn connect<T>(
mut tls: T,

Some files were not shown because too many files have changed in this diff Show More