## 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
## Problem
We don't have good observability for "stuck" getpage requests.
Resolves https://github.com/neondatabase/cloud/issues/23808.
## Summary of changes
Log a periodic warning (every 30 seconds) if GetPage request execution
is slow to complete, to aid in debugging stuck GetPage requests.
This does not cover response flushing (we have separate logging for
that), nor reading the request from the socket and batching it (expected
to be insignificant and not straightforward to handle with the current
protocol).
This costs 95 nanoseconds on the happy path when awaiting a
`tokio::task::yield_now()`:
```
warn_slow/enabled=false time: [45.716 ns 46.116 ns 46.687 ns]
warn_slow/enabled=true time: [141.53 ns 141.83 ns 142.18 ns]
```
## Problem
The interpreted SK <-> PS protocol does not guard against gaps (neither
does the Vanilla one, but that's beside the point).
## Summary of changes
Extend the protocol to include the start LSN of the PG WAL section from
which the records were interpreted.
Validation is enabled via a config flag on the pageserver and works as
follows:
**Case 1**: `raw_wal_start_lsn` is smaller than the requested LSN
There can't be gaps here, but we check that the shard received records
which it hasn't seen before.
**Case 2**: `raw_wal_start_lsn` is equal to the requested LSN
This is the happy case. No gap and nothing to check
**Case 3**: `raw_wal_start_lsn` is greater than the requested LSN
This is a gap.
To make Case 3 work I had to bend the protocol a bit.
We read record chunks of WAL which aren't record aligned and feed them
to the decoder.
The picture below shows a shard which subscribes at a position somewhere
within Record 2.
We already have a wal reader which is below that position so we wait to
catch up.
We read some wal in Read 1 (all of Record 1 and some of Record 2). The
new shard doesn't
need Record 1 (it has already processed it according to the starting
position), but we read
past it's starting position. When we do Read 2, we decode Record 2 and
ship it off to the shard,
but the starting position of Read 2 is greater than the starting
position the shard requested.
This looks like a gap.
To make it work, we extend the protocol to send an empty
`InterpretedWalRecords` to shards
if the WAL the records originated from ends the requested start
position. On the pageserver,
that just updates the tracking LSNs in memory (no-op really). This gives
us a workaround for
the fake gap.
As a drive by, make `InterpretedWalRecords::next_record_lsn` mandatory
in the application level definition.
It's always included.
Related: https://github.com/neondatabase/cloud/issues/23935
## Problem
Storage controller uses unsecure http for pageserver API.
Closes: https://github.com/neondatabase/cloud/issues/23734
Closes: https://github.com/neondatabase/cloud/issues/24091
## Summary of changes
- Add an optional `listen_https_port` field to storage controller's Node
state and its API (RegisterNode/ListNodes/etc).
- Allow updating `listen_https_port` on node registration to gradually
add https port for all nodes.
- Add `use_https_pageserver_api` CLI option to storage controller to
enable https.
- Pageserver doesn't support https for now and always reports
`https_port=None`. This will be addressed in follow-up PR.
ALTER SUBSCRIPTION requires AccessExclusive lock
which conflicts with iteration over pg_subscription when multiple
databases are present
and operations are applied concurrently.
Fix by explicitly locking pg_subscription
in the beginning of the transaction in each database.
## Problem
https://github.com/neondatabase/cloud/issues/24292
`pprof::symbolize()` used a regex to strip the Rust monomorphization
suffix from generic methods. However, the `backtrace` crate can do this
itself if formatted with the `:#` flag.
Also tighten up the code a bit.
## Problem
We've seen the previous default of 50 cause OOMs. Compacting many L0
layers at once now has limited benefit, since the cost is mostly linear
anyway. This is already being reduced to 20 in production settings.
## Summary of changes
Reduce `DEFAULT_COMPACTION_UPPER_LIMIT` to 20.
Once released, let's remove the config overrides.
## Problem
We lack an API for warming up attached locations based on the heatmap
contents.
This is problematic in two places:
1. If we manually migrate and cut over while the secondary is still cold
2. When we re-attach a previously offloaded tenant
## Summary of changes
https://github.com/neondatabase/neon/pull/10597 made heatmap generation
additive
across migrations, so we won't clobber it a after a cold migration. This
allows us to implement:
1. An endpoint for downloading all missing heatmap layers on the
pageserver:
`/v1/tenant/:tenant_shard_id/timeline/:timeline_id/download_heatmap_layers`.
Only one such operation per timeline is allowed at any given time. The
granularity is tenant shard.
2. An endpoint to the storage controller to trigger the downloads on the
pageserver:
`/v1/tenant/:tenant_shard_id/timeline/:timeline_id/download_heatmap_layers`.
This works both at
tenant and tenant shard level. If an unsharded tenant id is provided,
the operation is started on
all shards, otherwise only the specified shard.
3. A storcon cli command. Again, tenant and tenant-shard level
granularities are supported.
Cplane will call into storcon and trigger the downloads for all shards.
When we want to rescue a migration, we will use storcon cli targeting
the specific tenant shard.
Related: https://github.com/neondatabase/neon/issues/10541
Preparations for a successor of #10440:
* move `pull_timeline` to `safekeeper_api` and add it to
`SafekeeperClient`. we want to do `pull_timeline` on any creations that
we couldn't do initially.
* Add a `SafekeeperGeneration` type instead of relying on a type alias.
we want to maintain a safekeeper specific generation number now in the
storcon database. A separate type is important to make it impossible to
mix it up with the tenant's pageserver specific generation number. We
absolutely want to avoid that for correctness reasons. If someone mixes
up a safekeeper and pageserver id (both use the `NodeId` type), that's
bad but there is no wrong generations flying around.
part of #9011
I was looking into
https://github.com/neondatabase/serverless/issues/144, I recall previous
cases where proxy would trigger these prepared statements which would
conflict with other statements prepared by our client downstream.
Because of that, and also to aid in debugging, I've made sure all
prepared statements that proxy needs to make have specific names that
likely won't conflict and makes it clear in a error log if it's our
statements that are causing issues
## Problem
In #10707 some new fields were introduced in TimelineInfo.
I forgot that we do not only use TimelineInfo for encoding, but also
decoding when the storage controller calls into a pageserver, so this
broke some calls from controller to pageserver while in a mixed-version
state.
## Summary of changes
- Make new fields have default behavior so that they are optional
## Problem
Part of https://github.com/neondatabase/neon/issues/9516
## Summary of changes
This patch adds the support for storing reldir in the sparse keyspace.
All logic are guarded with the `rel_size_v2_enabled` flag, so if it's
set to false, the code path is exactly the same as what's currently in
prod.
Note that we did not persist the `rel_size_v2_enabled` flag and the
logic around it will be implemented in the next patch. (i.e., what if we
enabled it, restart the pageserver, and then it gets set to false? we
should still read from v2 using the rel_size_v2_migration_status in the
index_part). The persistence logic I'll implement in the next patch will
disallow switching from v2->v1 via config item.
I also refactored the metrics so that it can work with the new reldir
store. However, this metric is not correctly computed for reldirs (see
the comments) before. With the refactor, the value will be computed only
when we have an initial value for the reldir size. The refactor keeps
the incorrectness of the computation when there are more than 1
database.
For the tests, we currently run all the tests with v2, and I'll set it
to false and add some v2-specific tests before merging, probably also
v1->v2 migration tests.
---------
Signed-off-by: Alex Chi Z <chi@neon.tech>
# Summary
In
- https://github.com/neondatabase/neon/pull/10813
we added slow flush logging but it didn't log the TCP send & recv queue
length.
This PR adds that data to the log message.
I believe the implementation to be safe & correct right now, but it's
brittle and thus this PR should be reverted or improved upon once the
investigation is over.
Refs:
- stacked atop https://github.com/neondatabase/neon/pull/10813
- context:
https://neondb.slack.com/archives/C08DE6Q9C3B/p1739464533762049?thread_ts=1739462628.361019&cid=C08DE6Q9C3B
- improves https://github.com/neondatabase/neon/issues/10668
- part of https://github.com/neondatabase/cloud/issues/23515
# How It Works
The trouble is two-fold:
1. getting to the raw socket file descriptor through the many Rust types
that wrap it and
2. integrating with the `measure()` function
Rust wraps it in types to model file descriptor lifetimes and ownership,
and usually one can get access using `as_raw_fd()`.
However, we `split()` the stream and the resulting
[`tokio::io::WriteHalf`](https://docs.rs/tokio/latest/tokio/io/struct.WriteHalf.html)
.
Check the PR commit history for my attempts to do it.
My solution is to get the socket fd before we wrap it in our protocol
types, and to store that fd in the new `PostgresBackend::socket_fd`
field.
I believe it's safe because the lifetime of `PostgresBackend::socket_fd`
value == the lifetime of the `TcpStream` that wrap and store in
`PostgresBackend::framed`.
Specifically, the only place that close()s the socket is the `impl Drop
for TcpStream`.
I think the protocol stack calls `TcpStream::shutdown()`, but, that
doesn't `close()` the file descriptor underneath.
Regarding integration with the `measure()` function, the trouble is that
`flush_fut` is currently a generic `Future` type. So, we just pass in
the `socket_fd` as a separate argument.
A clean implementation would convert the `pgb_writer.flush()` to a named
future that provides an accessor for the socket fd while not being
polled.
I tried (see PR history), but failed to break through the `WriteHalf`.
# Testing
Tested locally by running
```
./target/debug/pagebench get-page-latest-lsn --num-clients=1000 --queue-depth=1000
```
in one terminal, waiting a bit, then
```
pkill -STOP pagebench
```
then wait for slow logs to show up in `pageserver.log`.
Pick one of the slow log message's port pairs, e.g., `127.0.0.1:39500`,
and then checking sockstat output
```
ss -ntp | grep '127.0.0.1:39500'
```
to ensure that send & recv queue size match those in the log message.
There is now a compute_ctl_config field in the response that currently
only contains a JSON Web Key set. compute_ctl currently doesn't do
anything with the keys, but will in the future.
The reasoning for the new field is due to the nature of empty computes.
When an empty compute is created, it does not have a tenant. A compute
spec is the primary means of communicating the details of an attached
tenant. In the empty compute state, there is no spec. Instead we wait
for the control plane to pass us one via /configure. If we were to
include the jwks field in the compute spec, we would have a partial
compute spec, which doesn't logically make sense.
Instead, we can have two means of passing settings to the compute:
- spec: tenant specific config details
- compute_ctl_config: compute specific settings
For instance, the JSON Web Key set passed to the compute is independent
of any tenant. It is a setting of the compute whether it is attached or
not.
Signed-off-by: Tristan Partin <tristan@neon.tech>
## Problem
Previously, when cutting over to cold secondary locations,
we would clobber the previous, good, heatmap with a cold one.
This is because heatmap generation used to include only resident layers.
Once this merges, we can add an endpoint which triggers full heatmap
hydration on attached locations to heal cold migrations.
## Summary of changes
With this patch, heatmap generation becomes additive. If we have a
heatmap from when this location was secondary, the new uploaded heatmap
will be the result of a reconciliation between the old one and the on
disk resident layers.
More concretely, when we have the previous heatmap:
1. Filter the previous heatmap and keep layers that are (a) present
in the current layer map, (b) visible, (c) not resident. Call this set
of layers `visible_non_resident`.
2. From the layer map, select all layers that are resident and visible.
Call this set of layers `resident`.
3. The new heatmap is the result of merging the two disjoint sets.
Related https://github.com/neondatabase/neon/issues/10541
## Problem
We expose `latest_gc_cutoff` in our API, and callers understandably were
using that to validate LSNs for branch creation. However, this is _not_
the true GC cutoff from a user's point of view: it's just the point at
which we last actually did GC. The actual cutoff used when validating
branch creations and page_service reads is the min() of latest_gc_cutoff
and the planned GC lsn in GcInfo.
Closes: https://github.com/neondatabase/neon/issues/10639
## Summary of changes
- Expose the more useful min() of GC cutoffs as `gc_cutoff_lsn` in the
API, so that the most obviously named field is really the one people
should use.
- Retain the ability to read the LSN at which GC was actually done, in
an `applied_gc_cutoff_lsn` field.
- Internally rename `latest_gc_cutoff_lsn` to `applied_gc_cutoff_lsn`
("latest" was a confusing name, as the value in GcInfo is more up to
date in terms of what a user experiences)
- Temporarily preserve the old `latest_gc_cutoff_lsn` field for compat
with control plane until we update it to use the new field.
---------
Co-authored-by: Arpad Müller <arpad-m@users.noreply.github.com>
## Problem
L0 compaction frequently gets starved out by other background tasks and
image/GC compaction. L0 compaction must be responsive to keep read
amplification under control.
Touches #10694.
Resolves#10689.
## Summary of changes
Use a separate semaphore for the L0-only compaction pass.
* Add a `CONCURRENT_L0_COMPACTION_TASKS` semaphore and
`BackgroundLoopKind::L0Compaction`.
* Add a setting `compaction_l0_semaphore` (default off via
`compaction_l0_first`).
* Use the L0 semaphore when doing an `OnlyL0Compaction` pass.
* Use the background semaphore when doing a regular compaction pass
(which includes an initial L0 pass).
* While waiting for the background semaphore, yield for L0 compaction if
triggered.
* Add `CompactFlags::NoYield` to disable L0 yielding, and set it for the
HTTP API route.
* Remove the old `use_compaction_semaphore` setting and
compaction-scoped semaphore.
* Remove the warning when waiting for a semaphore; it's noisy and we
have metrics.
## Problem
Image compaction can starve out L0 compaction if a tenant has several
timelines with L0 debt.
Touches #10694.
Requires #10740.
## Summary of changes
* Add an initial L0 compaction pass, in order of L0 count.
* Add a tenant option `compaction_l0_first` to control the L0 pass
(disabled by default).
* Add `CompactFlags::OnlyL0Compaction` to run an L0-only compaction
pass.
* Clean up the compaction iteration logic.
A later PR will use separate semaphores for the L0 and image compaction
passes to avoid cross-tenant L0 starvation. That PR will also make image
compaction yield if _any_ of the tenant's timelines have pending L0
compaction to further avoid starvation.
Avoids compiling the crate and its dependencies into binaries that don't
need them. Shrinks the compute_ctl binary from about 31MB to 28MB in the
release-line-debug-size-lto profile.
This changes the default value of the `timeline_offloading` pageserver
and tenant configs to true, now that offloading has been rolled out
without problems.
There is also a small fix in the tenant config merge function, where we
applied the `lazy_slru_download` value instead of `timeline_offloading`.
Related issue: https://github.com/neondatabase/cloud/issues/21353
## Problem
The compaction loop currently runs periodically, which can cause it to
wait for up to 20 seconds before starting L0 compaction by default.
Also, when we later separate the semaphores for L0 compaction and image
compaction, we want to give up waiting for the image compaction
semaphore if L0 compaction is needed on any timeline.
Touches #10694.
## Summary of changes
Notify the compaction loop when an L0 flush (on any timeline) exceeds
`compaction_threshold`.
Also do some opportunistic cleanups in the area.
## Problem
helps investigate https://github.com/neondatabase/neon/issues/10482
## Summary of changes
In debug mode and testing mode, we will record all files visited by a
read operation, and print it out when it errors.
---------
Signed-off-by: Alex Chi Z <chi@neon.tech>
## Problem
L0 compaction can get starved by other background tasks. It needs to be
responsive to avoid read amp blowing up during heavy write workloads.
Touches #10694.
## Summary of changes
Add a separate semaphore for compaction, configurable via
`use_compaction_semaphore` (disabled by default). This is primarily for
testing in staging; it needs further work (in particular to split
image/L0 compaction jobs) before it can be enabled.
## Problem
Following #10641, let's add a few critical errors.
Resolves#10094.
## Summary of changes
Adds the following critical errors:
* WAL sender read/decode failure.
* WAL record ingestion failure.
* WAL redo failure.
* Missing key during compaction.
We don't add an error for missing keys during GetPage requests, since
we've seen a handful of these in production recently, and the cause is
still unclear (most likely a benign race).
## Problem
The local filesystem backend for remote storage doesn't set a
concurrency limit. While it can't/won't enforce a concurrency limit
itself, this also bounds the upload queue concurrency. Some tests create
thousands of uploads, which slows down the quadratic scheduling of the
upload queue, and there is no point spawning that many Tokio tasks.
Resolves#10409.
## Summary of changes
Set a concurrency limit of 100 for the LocalFS backend.
Before: `test_layer_map[release-pg17].test_query: 68.338 s`
After: `test_layer_map[release-pg17].test_query: 5.209 s`
Given a remote extensions manifest of the following:
```json
{
"public_extensions": [],
"custom_extensions": null,
"library_index": {
"pg_search": "pg_search"
},
"extension_data": {
"pg_search": {
"control_data": {
"pg_search.control": "comment = 'pg_search: Full text search for PostgreSQL using BM25'\ndefault_version = '0.14.1'\nmodule_pathname = '$libdir/pg_search'\nrelocatable = false\nsuperuser = true\nschema = paradedb\ntrusted = true\n"
},
"archive_path": "13117844657/v14/extensions/pg_search.tar.zst"
}
}
}
```
We were allowing a compute to install a remote extension that wasn't
listed in either public_extensions or custom_extensions.
Signed-off-by: Tristan Partin <tristan@neon.tech>
## Problem
Image layer generation could block L0 compactions for a long time.
## Summary of changes
* Refactored the return value of `create_image_layers_for_*` functions
to make it self-explainable.
* Preempt image layer generation in `Try` mode if L0 piles up.
Note that we might potentially run into a state that only the beginning
part of the keyspace gets image coverage. In that case, we either need
to implement something to prioritize some keyspaces with image coverage,
or tune the image_creation_threshold to ensure that the frequency of
image creation could keep up with L0 compaction.
---------
Signed-off-by: Alex Chi Z <chi@neon.tech>
Co-authored-by: Erik Grinaker <erik@neon.tech>
## Problem
We don't currently have good alerts for critical errors, e.g. data
loss/corruption.
Touches #10094.
## Summary of changes
Add a `critical!` macro and corresponding
`libmetrics_tracing_event_count{level="critical"}` metric. This will:
* Emit an `ERROR` log message with prefix `"CRITICAL:"` and a backtrace.
* Increment `libmetrics_tracing_event_count{level="critical"}`, and
indirectly `level="error"`.
* Trigger a pageable alert (via the metric above).
* In debug builds, panic the process.
I'll add uses of the macro separately.
Logging errors with the debug format specifier causes multi-line errors,
which are sometimes a pain to deal with. Instead, we should use anyhow's
alternate display format, which shows the same information on a single
line.
Also adjusted a couple of error messages that were stale.
Fixesneondatabase/cloud#14710.
## Problem
The approach of having CancelMap as an in-memory structure increases
code complexity,
as well as putting additional load for Redis streams.
## Summary of changes
- Implement a set of KV ops for Redis client;
- Remove cancel notifications code;
- Send KV ops over the bounded channel to the handling background task
for removing and adding the cancel keys.
Closes#9660
## Problem
Follow-up of the incident, we should not use the same bound on
lower/upper limit of compaction files. This patch adds an upper bound
limit, which is set to 50 for now.
## Summary of changes
Add `compaction_upper_limit`.
---------
Signed-off-by: Alex Chi Z <chi@neon.tech>
Co-authored-by: Christian Schwarz <christian@neon.tech>
## Problem
We've seen the ingest connection manager get stuck shortly after a
migration.
## Summary of changes
A speculative mitigation is to use the same mechanism as get page
requests for kicking LSN ingest. The connection manager monitors
LSN waits and queries the broker if no updates are received for the
timeline.
Closes https://github.com/neondatabase/neon/issues/10351
## Problem
We need a setting to disable the flush upload wait, to test L0 flush
backpressure in staging.
## Summary of changes
Add `l0_flush_wait_upload` setting.
Only a few things that needed updating:
- async_trait was removed
- Message::Text takes a Utf8Bytes object instead of a String
Signed-off-by: Tristan Partin <tristan@neon.tech>
Co-authored-by: Conrad Ludgate <connor@neon.tech>
## Problem
We were logging a warning after a single request timeout, while listing
objects.
Closes: https://github.com/neondatabase/neon/issues/10166
## Summary of changes
- These timeouts are a pretty normal part of life, so back it off to
only log a warning after two in a row.
## Problem
There is no direct backpressure for compaction and L0 read
amplification. This allows a large buildup of compaction debt and read
amplification.
Resolves#5415.
Requires #10402.
## Summary of changes
Delay layer flushes based on the number of level 0 delta layers:
* `l0_flush_delay_threshold`: delay flushes such that they take 2x as
long (default `2 * compaction_threshold`).
* `l0_flush_stall_threshold`: stall flushes until level 0 delta layers
drop below threshold (default `4 * compaction_threshold`).
If either threshold is reached, ephemeral layer rolls also synchronously
wait for layer flushes to propagate this backpressure up into WAL
ingestion. This will bound the number of frozen layers to 1 once
backpressure kicks in, since all other frozen layers must flush before
the rolled layer.
## Analysis
This will significantly change the compute backpressure characteristics.
Recall the three compute backpressure knobs:
* `max_replication_write_lag`: 500 MB (based on Pageserver
`last_received_lsn`).
* `max_replication_flush_lag`: 10 GB (based on Pageserver
`disk_consistent_lsn`).
* `max_replication_apply_lag`: disabled (based on Pageserver
`remote_consistent_lsn`).
Previously, the Pageserver would keep ingesting WAL and build up
ephemeral layers and L0 layers until the compute hit
`max_replication_flush_lag` at 10 GB and began backpressuring. Now, once
we delay/stall WAL ingestion, the compute will begin backpressuring
after `max_replication_write_lag`, i.e. 500 MB. This is probably a good
thing (we're not building up a ton of compaction debt), but we should
consider tuning these settings.
`max_replication_flush_lag` probably doesn't serve a purpose anymore,
and we should consider removing it.
Furthermore, the removal of the upload barrier in #10402 will mean that
we no longer backpressure flushes based on S3 uploads, since
`max_replication_apply_lag` is disabled. We should consider enabling
this as well.
### When and what do we compact?
Default compaction settings:
* `compaction_threshold`: 10 L0 delta layers.
* `compaction_period`: 20 seconds (between each compaction loop check).
* `checkpoint_distance`: 256 MB (size of L0 delta layers).
* `l0_flush_delay_threshold`: 20 L0 delta layers.
* `l0_flush_stall_threshold`: 40 L0 delta layers.
Compaction characteristics:
* Minimum compaction volume: 10 layers * 256 MB = 2.5 GB.
* Additional compaction volume (assuming 128 MB/s WAL): 128 MB/s * 20
seconds = 2.5 GB (10 L0 layers).
* Required compaction bandwidth: 5.0 GB / 20 seconds = 256 MB/s.
### When do we hit `max_replication_write_lag`?
Depending on how fast compaction and flushes happens, the compute will
backpressure somewhere between `l0_flush_delay_threshold` or
`l0_flush_stall_threshold` + `max_replication_write_lag`.
* Minimum compute backpressure lag: 20 layers * 256 MB + 500 MB = 5.6 GB
* Maximum compute backpressure lag: 40 layers * 256 MB + 500 MB = 10.0
GB
This seems like a reasonable range to me.
Drop logical replication subscribers
before compute starts on a non-main branch.
Add new compute_ctl spec flag: drop_subscriptions_before_start
If it is set, drop all the subscriptions from the compute node
before it starts.
To avoid race on compute start, use new GUC
neon.disable_logical_replication_subscribers
to temporarily disable logical replication workers until we drop the
subscriptions.
Ensure that we drop subscriptions exactly once when endpoint starts on a
new branch.
It is essential, because otherwise, we may drop not only inherited, but
newly created subscriptions.
We cannot rely only on spec.drop_subscriptions_before_start flag,
because if for some reason compute restarts inside VM,
it will start again with the same spec and flag value.
To handle this, we save the fact of the operation in the database
in the neon.drop_subscriptions_done table.
If the table does not exist, we assume that the operation was never
performed, so we must do it.
If table exists, we check if the operation was performed on the current
timeline.
fixes: https://github.com/neondatabase/neon/issues/8790
## Refs
- Epic: https://github.com/neondatabase/neon/issues/9378
Co-authored-by: Vlad Lazar <vlad@neon.tech>
Co-authored-by: Christian Schwarz <christian@neon.tech>
## Problem
The read path does its IOs sequentially.
This means that if N values need to be read to reconstruct a page,
we will do N IOs and getpage latency is `O(N*IoLatency)`.
## Solution
With this PR we gain the ability to issue IO concurrently within one
layer visit **and** to move on to the next layer without waiting for IOs
from the previous visit to complete.
This is an evolved version of the work done at the Lisbon hackathon,
cf https://github.com/neondatabase/neon/pull/9002.
## Design
### `will_init` now sourced from disk btree index keys
On the algorithmic level, the only change is that the
`get_values_reconstruct_data`
now sources `will_init` from the disk btree index key (which is
PS-page_cache'd), instead
of from the `Value`, which is only available after the IO completes.
### Concurrent IOs, Submission & Completion
To separate IO submission from waiting for its completion, while
simultaneously
feature-gating the change, we introduce the notion of an `IoConcurrency`
struct
through which IO futures are "spawned".
An IO is an opaque future, and waiting for completions is handled
through
`tokio::sync::oneshot` channels.
The oneshot Receiver's take the place of the `img` and `records` fields
inside `VectoredValueReconstructState`.
When we're done visiting all the layers and submitting all the IOs along
the way
we concurrently `collect_pending_ios` for each value, which means
for each value there is a future that awaits all the oneshot receivers
and then calls into walredo to reconstruct the page image.
Walredo is now invoked concurrently for each value instead of
sequentially.
Walredo itself remains unchanged.
The spawned IO futures are driven to completion by a sidecar tokio task
that
is separate from the task that performs all the layer visiting and
spawning of IOs.
That tasks receives the IO futures via an unbounded mpsc channel and
drives them to completion inside a `FuturedUnordered`.
(The behavior from before this PR is available through
`IoConcurrency::Sequential`,
which awaits the IO futures in place, without "spawning" or "submitting"
them
anywhere.)
#### Alternatives Explored
A few words on the rationale behind having a sidecar *task* and what
alternatives were considered.
One option is to queue up all IO futures in a FuturesUnordered that is
polled
the first time when we `collect_pending_ios`.
Firstly, the IO futures are opaque, compiler-generated futures that need
to be polled at least once to submit their IO. "At least once" because
tokio-epoll-uring may not be able to submit the IO to the kernel on
first
poll right away.
Second, there are deadlocks if we don't drive the IO futures to
completion
independently of the spawning task.
The reason is that both the IO futures and the spawning task may hold
some
_and_ try to acquire _more_ shared limited resources.
For example, both spawning task and IO future may try to acquire
* a VirtualFile file descriptor cache slot async mutex (observed during
impl)
* a tokio-epoll-uring submission slot (observed during impl)
* a PageCache slot (currently this is not the case but we may move more
code into the IO futures in the future)
Another option is to spawn a short-lived `tokio::task` for each IO
future.
We implemented and benchmarked it during development, but found little
throughput improvement and moderate mean & tail latency degradation.
Concerns about pressure on the tokio scheduler made us discard this
variant.
The sidecar task could be obsoleted if the IOs were not arbitrary code
but a well-defined struct.
However,
1. the opaque futures approach taken in this PR allows leaving the
existing
code unchanged, which
2. allows us to implement the `IoConcurrency::Sequential` mode for
feature-gating
the change.
Once the new mode sidecar task implementation is rolled out everywhere,
and `::Sequential` removed, we can think about a descriptive submission
& completion interface.
The problems around deadlocks pointed out earlier will need to be solved
then.
For example, we could eliminate VirtualFile file descriptor cache and
tokio-epoll-uring slots.
The latter has been drafted in
https://github.com/neondatabase/tokio-epoll-uring/pull/63.
See the lengthy doc comment on `spawn_io()` for more details.
### Error handling
There are two error classes during reconstruct data retrieval:
* traversal errors: index lookup, move to next layer, and the like
* value read IO errors
A traversal error fails the entire get_vectored request, as before this
PR.
A value read error only fails that value.
In any case, we preserve the existing behavior that once
`get_vectored` returns, all IOs are done. Panics and failing
to poll `get_vectored` to completion will leave the IOs dangling,
which is safe but shouldn't happen, and so, a rate-limited
log statement will be emitted at warning level.
There is a doc comment on `collect_pending_ios` giving more code-level
details and rationale.
### Feature Gating
The new behavior is opt-in via pageserver config.
The `Sequential` mode is the default.
The only significant change in `Sequential` mode compared to before
this PR is the buffering of results in the `oneshot`s.
## Code-Level Changes
Prep work:
* Make `GateGuard` clonable.
Core Feature:
* Traversal code: track `will_init` in `BlobMeta` and source it from
the Delta/Image/InMemory layer index, instead of determining `will_init`
after we've read the value. This avoids having to read the value to
determine whether traversal can stop.
* Introduce `IoConcurrency` & its sidecar task.
* `IoConcurrency` is the clonable handle.
* It connects to the sidecar task via an `mpsc`.
* Plumb through `IoConcurrency` from high level code to the
individual layer implementations' `get_values_reconstruct_data`.
We piggy-back on the `ValuesReconstructState` for this.
* The sidecar task should be long-lived, so, `IoConcurrency` needs
to be rooted up "high" in the call stack.
* Roots as of this PR:
* `page_service`: outside of pagestream loop
* `create_image_layers`: when it is called
* `basebackup`(only auxfiles + replorigin + SLRU segments)
* Code with no roots that uses `IoConcurrency::sequential`
* any `Timeline::get` call
* `collect_keyspace` is a good example
* follow-up: https://github.com/neondatabase/neon/issues/10460
* `TimelineAdaptor` code used by the compaction simulator, unused in
practive
* `ingest_xlog_dbase_create`
* Transform Delta/Image/InMemoryLayer to
* do their values IO in a distinct `async {}` block
* extend the residence of the Delta/Image layer until the IO is done
* buffer their results in a `oneshot` channel instead of straight
in `ValuesReconstructState`
* the `oneshot` channel is wrapped in `OnDiskValueIo` /
`OnDiskValueIoWaiter`
types that aid in expressiveness and are used to keep track of
in-flight IOs so we can print warnings if we leave them dangling.
* Change `ValuesReconstructState` to hold the receiving end of the
`oneshot` channel aka `OnDiskValueIoWaiter`.
* Change `get_vectored_impl` to `collect_pending_ios` and issue walredo
concurrently, in a `FuturesUnordered`.
Testing / Benchmarking:
* Support queue-depth in pagebench for manual benchmarkinng.
* Add test suite support for setting concurrency mode ps config
field via a) an env var and b) via NeonEnvBuilder.
* Hacky helper to have sidecar-based IoConcurrency in tests.
This will be cleaned up later.
More benchmarking will happen post-merge in nightly benchmarks, plus in
staging/pre-prod.
Some intermediate helpers for manual benchmarking have been preserved in
https://github.com/neondatabase/neon/pull/10466 and will be landed in
later PRs.
(L0 layer stack generator!)
Drive-By:
* test suite actually didn't enable batching by default because
`config.compatibility_neon_binpath` is always Truthy in our CI
environment
=> https://neondb.slack.com/archives/C059ZC138NR/p1737490501941309
* initial logical size calculation wasn't always polled to completion,
which was
surfaced through the added WARN logs emitted when dropping a
`ValuesReconstructState` that still has inflight IOs.
* remove the timing histograms
`pageserver_getpage_get_reconstruct_data_seconds`
and `pageserver_getpage_reconstruct_seconds` because with planning,
value read
IO, and walredo happening concurrently, one can no longer attribute
latency
to any one of them; we'll revisit this when Vlad's work on
tracing/sampling
through RequestContext lands.
* remove code related to `get_cached_lsn()`.
The logic around this has been dead at runtime for a long time,
ever since the removal of the materialized page cache in #8105.
## Testing
Unit tests use the sidecar task by default and run both modes in CI.
Python regression tests and benchmarks also use the sidecar task by
default.
We'll test more in staging and possibly preprod.
# Future Work
Please refer to the parent epic for the full plan.
The next step will be to fold the plumbing of IoConcurrency
into RequestContext so that the function signatures get cleaned up.
Once `Sequential` isn't used anymore, we can take the next
big leap which is replacing the opaque IOs with structs
that have well-defined semantics.
---------
Co-authored-by: Christian Schwarz <christian@neon.tech>
# Refs
- extracted from https://github.com/neondatabase/neon/pull/9353
# Problem
Before this PR, when task_mgr shutdown is signalled, e.g. during
pageserver shutdown or Tenant shutdown, initial logical size calculation
stops polling and drops the future that represents the calculation.
This is against the current policy that we poll all futures to
completion.
This became apparent during development of concurrent IO which warns if
we drop a `Timeline::get_vectored` future that still has in-flight IOs.
We may revise the policy in the future, but, right now initial logical
size calculation is the only part of the codebase that doesn't adhere to
the policy, so let's fix it.
## Code Changes
- make sensitive exclusively to `Timeline::cancel`
- This should be sufficient for all cases of shutdowns; the sensitivity
to task_mgr shutdown is unnecessary.
- this broke the various cancel tests in `test_timeline_size.py`, e.g.,
`test_timeline_initial_logical_size_calculation_cancellation`
- the tests would time out because the await point was not sensitive to
cancellation
- to fix this, refactor `pausable_failpoint` so that it accepts a
cancellation token
- side note: we _really_ should write our own failpoint library; maybe
after we get heap-allocated RequestContext, we can plumb failpoints
through there.
## Problem
part of https://github.com/neondatabase/neon/issues/9114
The automatic trigger is already implemented at
https://github.com/neondatabase/neon/pull/10221 but I need to write some
tests and finish my experiments in staging before I can merge it with
confidence. Given that I have some other patches that will modify the
config items, I'd like to get the config items merged first to reduce
conflicts.
## Summary of changes
* add `l2_lsn` to index_part.json -- below that LSN, data have been
processed by gc-compaction
* add a set of gc-compaction auto trigger control items into the config
---------
Signed-off-by: Alex Chi Z <chi@neon.tech>
Add an endpoint to obtain the utilization of a safekeeper. Future
changes to the storage controller can use this endpoint to find the most
suitable safekeepers for newly created timelines, analogously to how
it's done for pageservers already.
Initially we just want to assign by timeline count, then we can iterate
from there.
Part of https://github.com/neondatabase/neon/issues/9011
