Stacks on:
- https://github.com/neondatabase/neon/pull/7165
Fixes while working on background optimization of scheduling after a
split:
- When a tenant has secondary locations, we weren't detaching the parent
shards' secondary locations when doing a split
- When a reconciler detaches a location, it was feeding back a
locationconf with `Detached` mode in its `observed` object, whereas it
should omit that location. This could cause the background reconcile
task to keep kicking off no-op reconcilers forever (harmless but
annoying).
- During shard split, we were scheduling secondary locations for the
child shards, but no reconcile was run for these until the next time the
background reconcile task ran. Creating these ASAP is useful, because
they'll be used shortly after a shard split as the destination locations
for migrating the new shards to different nodes.
## Problem
Storage controller had basically no metrics.
## Summary of changes
1. Migrate the existing metrics to use Conrad's
[`measured`](https://docs.rs/measured/0.0.14/measured/) crate.
2. Add metrics for incoming http requests
3. Add metrics for outgoing http requests to the pageserver
4. Add metrics for outgoing pass through requests to the pageserver
5. Add metrics for database queries
Note that the metrics response for the attachment service does not use
chunked encoding like the rest of the metrics endpoints. Conrad has
kindly extended the crate such that it can now be done. Let's leave it
for a follow-up since the payload shouldn't be that big at this point.
Fixes https://github.com/neondatabase/neon/issues/6875
## Problem
The current implementation of struct Layer supports canceled read
requests, but those will leave the internal state such that a following
`Layer::keep_resident` call will need to repair the state. In
pathological cases seen during generation numbers resetting in staging
or with too many in-progress on-demand downloads, this repair activity
will need to wait for the download to complete, which stalls disk
usage-based eviction. Similar stalls have been observed in staging near
disk-full situations, where downloads failed because the disk was full.
Fixes#6028 or the "layer is present on filesystem but not evictable"
problems by:
1. not canceling pending evictions by a canceled
`LayerInner::get_or_maybe_download`
2. completing post-download initialization of the `LayerInner::inner`
from the download task
Not canceling evictions above case (1) and always initializing (2) lead
to plain `LayerInner::inner` always having the up-to-date information,
which leads to the old `Layer::keep_resident` never having to wait for
downloads to complete. Finally, the `Layer::keep_resident` is replaced
with `Layer::is_likely_resident`. These fix#7145.
## Summary of changes
- add a new test showing that a canceled get_or_maybe_download should
not cancel the eviction
- switch to using a `watch` internally rather than a `broadcast` to
avoid hanging eviction while a download is ongoing
- doc changes for new semantics and cleanup
- fix `Layer::keep_resident` to use just `self.0.inner.get()` as truth
as `Layer::is_likely_resident`
- remove `LayerInner::wanted_evicted` boolean as no longer needed
Builds upon: #7185. Cc: #5331.
Before this PR, cancellation for `LayerInner::get_or_maybe_download`
could occur so that we have downloaded the layer file in the filesystem,
but because of the cancellation chance, we have not set the internal
`LayerInner::inner` or initialized the state. With the detached init
support introduced in #7135 and in place in #7152, we can now initialize
the internal state after successfully downloading in the spawned task.
The next PR will fix the remaining problems that this PR leaves:
- `Layer::keep_resident` is still used because
- `Layer::get_or_maybe_download` always cancels an eviction, even when
canceled
Split off from #7030. Stacked on top of #7152. Cc: #5331.
- Enable debug logs for this test
- Add some debug logging detail in downloader.rs
- Add an info-level message in scheduler.rs that makes it obvious if a
command is waiting for an existing task rather than spawning a new one.
The second part of work towards fixing `Layer::keep_resident` so that it
does not need to repair the internal state. #7135 added a nicer API for
initialization. This PR uses it to remove a few indentation levels and
the loop construction. The next PR #7175 will use the refactorings done
in this PR, and always initialize the internal state after a download.
Cc: #5331
Since #6115 with more often used get_value_reconstruct_data and friends,
we should not have needless INFO level span creation near hot paths. In
our prod configuration, INFO spans are always created, but in practice,
very rarely anything at INFO level is logged underneath.
`ResidentLayer::load_keys` is only used during compaction so it is not
that hot, but this aligns the access paths and their span usage.
PR changes the span level to debug to align with others, and adds the
layer name to the error which was missing.
Split off from #7030.
Add shard_number to PageserverFeedback and parse it on the compute side.
When compute receives a new ps_feedback, it calculates min LSNs among
feedbacks from all shards, and uses those LSNs for backpressure.
Add `test_sharding_backpressure` to verify that backpressure slows down
compute to wait for the slowest shard.
Manual testing of the changes in #7160 revealed that, if the
thread-local destructor ever runs (it apparently doesn't in our test
suite runs, otherwise #7160 would not have auto-merged), we can
encounter an `abort()` due to a double-panic in the tracing code.
This github comment here contains the stack trace:
https://github.com/neondatabase/neon/pull/7160#issuecomment-2003778176
This PR reverts #7160 and uses a atomic counter to identify the
thread-local in log messages, instead of the memory address of the
thread local, which may be re-used.
with `immediate_gc` the span only covered the `gc_iteration`, make it
cover the whole needless spawned task, which also does waiting for layer
drops and stray logging in tests.
also clarify some comments while we are here.
Fixes: #6910
The PR #7141 added log message
```
ThreadLocalState is being dropped and id might be re-used in the future
```
which was supposed to be emitted when the thread-local is destroyed.
Instead, it was emitted on _each_ call to `thread_local_system()`,
ie.., on each tokio-epoll-uring operation.
Testing
-------
Reproduced the issue locally and verified that this PR fixes the issue.
## Problem
Followup to https://github.com/neondatabase/neon/pull/6725
In that PR, code for purging local files from a tenant shard was
duplicated.
## Summary of changes
- Refactor detach code into TenantManager
- `spawn_background_purge` method can now be common between detach and
split operations
refs https://github.com/neondatabase/neon/issues/7136
Problem
-------
Before this PR, we were using
`tokio_epoll_uring::thread_local_system()`,
which panics on tokio_epoll_uring::System::launch() failure
As we've learned in [the
past](https://github.com/neondatabase/neon/issues/6373#issuecomment-1905814391),
some older Linux kernels account io_uring instances as locked memory.
And while we've raised the limit in prod considerably, we did hit it
once on 2024-03-11 16:30 UTC.
That was after we enabled tokio-epoll-uring fleet-wide, but before
we had shipped release-5090 (c6ed86d3d0)
which did away with the last mass-creation of tokio-epoll-uring
instances as per
commit 3da410c8fe
Author: Christian Schwarz <christian@neon.tech>
Date: Tue Mar 5 10:03:54 2024 +0100
tokio-epoll-uring: use it on the layer-creating code paths (#6378)
Nonetheless, it highlighted that panicking in this situation is probably
not ideal, as it can leave the pageserver process in a semi-broken
state.
Further, due to low sampling rate of Prometheus metrics, we don't know
much about the circumstances of this failure instance.
Solution
--------
This PR implements a custom thread_local_system() that is
pageserver-aware
and will do the following on failure:
- dump relevant stats to `tracing!`, hopefully they will be useful to
understand the circumstances better
- if it's the locked memory failure (or any other ENOMEM): abort() the
process
- if it's ENOMEM, retry with exponential back-off, capped at 3s.
- add metric counters so we can create an alert
This makes sense in the production environment where we know that
_usually_, there's ample locked memory allowance available, and we know
the failure rate is rare.
## Problem
The existing secondary download API relied on the caller to wait as long
as it took to complete -- for large shards that could be a long time, so
typical clients that might have a baked-in ~30s timeout would have a
problem.
## Summary of changes
- Take a `wait_ms` query parameter to instruct the pageserver how long
to wait: if the download isn't complete in this duration, then 201 is
returned instead of 200.
- For both 200 and 201 responses, include response body describing
download progress, in terms of layers and bytes. This is sufficient for
the caller to track how much data is being transferred and log/present
that status.
- In storage controller live migrations, use this API to apply a much
longer outer timeout, with smaller individual per-request timeouts, and
log the progress of the downloads.
- Add a test that injects layer download delays to exercise the new
behavior
# Problem
On-demand downloads are still using `tokio::fs`, which we know is
inefficient.
# Changes
- Add `pagebench ondemand-download-churn` to quantify on-demand download
throughput
- Requires dumping layer map, which required making `history_buffer`
impl `Deserialize`
- Implement an equivalent of `tokio::io::copy_buf` for owned buffers =>
`owned_buffers_io` module and children.
- Make layer file download sensitive to `io_engine::get()`, using
VirtualFile + above copy loop
- For this, I had to move some code into the `retry_download`, e.g.,
`sync_all()` call.
Drive-by:
- fix missing escaping in `scripts/ps_ec2_setup_instance_store`
- if we failed in retry_download to create a file, we'd try to remove
it, encounter `NotFound`, and `abort()` the process using
`on_fatal_io_error`. This PR adds treats `NotFound` as a success.
# Testing
Functional
- The copy loop is generic & unit tested.
Performance
- Used the `ondemand-download-churn` benchmark to manually test against
real S3.
- Results (public Notion page):
https://neondatabase.notion.site/Benchmarking-tokio-epoll-uring-on-demand-downloads-2024-04-15-newer-code-03c0fdc475c54492b44d9627b6e4e710?pvs=4
- Performance is equivalent at low concurrency. Jumpier situation at
high concurrency, but, still less CPU / throughput with
tokio-epoll-uring.
- It’s a win.
# Future Work
Turn the manual performance testing described in the above results
document into a performance regression test:
https://github.com/neondatabase/neon/issues/7146
## Problem
Tenant deletion had a couple of TODOs where we weren't using proper
cancellation tokens that would have aborted the deletions during process
shutdown.
## Summary of changes
- Refactor enough that deletion/shutdown code has access to the
TenantManager's cancellation toke
- Use that cancellation token in tenant deletion instead of dummy
tokens.
fixes https://github.com/neondatabase/neon/issues/7116
Changes:
- refactor PageServerConfigBuilder: support not-set values
- implement runtime feature test
- use runtime feature test to determine `virtual_file_io_engine` if not
explicitly configured in the config
- log the effective engine at startup
- drive-by: improve assertion messages in `test_pageserver_init_node_id`
This needed a tiny bit of tokio-epoll-uring work, hence bumping it.
Changelog:
```
git log --no-decorate --oneline --reverse 868d2c42b5d54ca82fead6e8f2f233b69a540d3e..342ddd197a060a8354e8f11f4d12994419fff939
c7a74c6 Bump mio from 0.8.8 to 0.8.11
4df3466 Bump mio from 0.8.8 to 0.8.11 (#47)
342ddd1 lifecycle: expose `LaunchResult` enum (#49)
```
## Problem
See:
- https://github.com/neondatabase/neon/issues/6374
## Summary of changes
Whereas previously we calculated synthetic size from the gc_horizon or
the pitr_interval (whichever is the lower LSN), now we ignore gc_horizon
and exclusively start from the `pitr_interval`. This is a more generous
calculation for billing, where we do not charge users for data retained
due to gc_horizon.
Split off from #7030:
- each early exit is counted as canceled init, even though it most
likely was just `LayerInner::keep_resident` doing the no-download repair
check
- `downloaded_after` could had been accounted for multiple times, and
also when repairing to match on-disk state
Cc: #5331
Switched the order; doing https://github.com/neondatabase/neon/pull/6139
first then can remove uninit marker after.
## Problem
Previously, existence of a timeline directory was treated as evidence of
the timeline's logical existence. That is no longer the case since we
treat remote storage as the source of truth on each startup: we can
therefore do without this mark file.
The mark file had also been used as a pseudo-lock to guard against
concurrent creations of the same TimelineId -- now that persistence is
no longer required, this is a bit unwieldy.
In #6139 the `Tenant::timelines_creating` was added to protect against
concurrent creations on the same TimelineId, making the uninit mark file
entirely redundant.
## Summary of changes
- Code that writes & reads mark file is removed
- Some nearby `pub` definitions are amended to `pub(crate)`
- `test_duplicate_creation` is added to demonstrate that mutual
exclusion of creations still works.
## Problem
If a pageserver was offline when the storage controller started, there
was no mechanism to update the
storage controller state when the pageserver becomes active.
## Summary of changes
* Add a heartbeater module. The heartbeater must be driven by an
external loop.
* Integrate the heartbeater into the service.
- Extend the types used by the service and scheduler to keep track of a
nodes' utilisation score.
- Add a background loop to drive the heartbeater and update the state
based on the deltas it generated
- Do an initial round of heartbeats at start-up
## Problem
Shard splits worked, but weren't safe against failures (e.g. node crash
during split) yet.
Related: #6676
## Summary of changes
- Introduce async rwlocks at the scope of Tenant and Node:
- exclusive tenant lock is used to protect splits
- exclusive node lock is used to protect new reconciliation process that
happens when setting node active
- exclusive locks used in both cases when doing persistent updates (e.g.
node scheduling conf) where the update to DB & in-memory state needs to
be atomic.
- Add failpoints to shard splitting in control plane and pageserver
code.
- Implement error handling in control plane for shard splits: this
detaches child chards and ensures parent shards are re-attached.
- Crash-safety for storage controller restarts requires little effort:
we already reconcile with nodes over a storage controller restart, so as
long as we reset any incomplete splits in the DB on restart (added in
this PR), things are implicitly cleaned up.
- Implement reconciliation with offline nodes before they transition to
active:
- (in this context reconciliation means something like
startup_reconcile, not literally the Reconciler)
- This covers cases where split abort cannot reach a node to clean it
up: the cleanup will eventually happen when the node is marked active,
as part of reconciliation.
- This also covers the case where a node was unavailable when the
storage controller started, but becomes available later: previously this
allowed it to skip the startup reconcile.
- Storage controller now terminates on panics. We only use panics for
true "should never happen" assertions, and these cases can leave us in
an un-usable state if we keep running (e.g. panicking in a shard split).
In the unlikely event that we get into a crashloop as a result, we'll
rely on kubernetes to back us off.
- Add `test_sharding_split_failures` which exercises a variety of
failure cases during shard split.
## Problem
Before this PR, `Timeline::get_vectored` would be throttled twice if the
sequential option was enabled or if validation was enabled.
Also, `pageserver_get_vectored_seconds` included the time spent in the
throttle, which turns out to be undesirable for what we use that metric
for.
## Summary of changes
Double-throttle:
* Add `Timeline::get0` method which is unthrottled.
* Use that method from within the `Timeline::get_vectored` code path.
Metric:
* return throttled time from `throttle()` method
* deduct the value from the observed time
* globally rate-limited logging of duration subtraction errors, like in
all other places that do the throttled-time deduction from observations
The `tenant_id` in `TenantLocationConfigRequest` in the
`location_config` endpoint was only used in the storage
controller/attachment service, and there it was only used for assertions
and the creation part.
Currently, the flushing operation could flush multiple frozen layers to
the disk and store the aggregate time in the histogram. The result is a
bimodal distribution with short and over 1000-second flushes. Change it
so that we record how long one layer flush takes.
## Summary
- Currently we can set stripe size at tenant creation, but it doesn't
mean anything until we have multiple shards
- When onboarding an existing tenant, it will always get a default shard
stripe size, so we would like to be able to pick the actual stripe size
at the point we split.
## Why do this inline with a split?
The alternative to this change would be to have a separate endpoint on
the storage controller for setting the stripe size on a tenant, and only
permit writes to that endpoint when the tenant has only a single shard.
That would work, but be a little bit more work for a client, and not
appreciably simpler (instead of having a special argument to the split
functions, we'd have a special separate endpoint, and a requirement that
the controller must sync its config down to the pageserver before
calling the split API). Either approach would work, but this one feels a
bit more robust end-to-end: the split API is the _very last moment_ that
the stripe size is mutable, so if we aim to set it before splitting, it
makes sense to do it as part of the same operation.
## Summary of changes
The problem it fixes is when `request_lsn` is `u64::MAX-1` the
`cont_lsn` becomes `u64::MAX` which is the same as `prev_lsn` which
stops the loop.
Closes https://github.com/neondatabase/neon/issues/6812
## Problem
Currently we manually register nodes with the storage controller, and
use a script during deploy to register with the cloud control plane.
Rather than extend that script further, nodes should just register on
startup.
## Summary of changes
- Extend the re-attach request to include an optional
NodeRegisterRequest
- If the `register` field is set, handle it like a normal node
registration before executing the normal re-attach work.
- Update tests/neon_local that used to rely on doing an explicit
register step that could be enabled/disabled.
---------
Co-authored-by: Christian Schwarz <christian@neon.tech>
This is a follow-up to #7051 where `LayerInner::drop` and
`LayerInner::evict_blocking` were not noticed to require a gate before
the file deletion. The lack of entering a gate opens up a similar
possibility of deleting a layer file which a newer Timeline instance has
already checked out to be resident in a similar case as #7051.
Tenant::shutdown or Timeline::shutdown completes and becomes externally
observable before the corresponding Tenant/Timeline object is dropped.
For example, after observing a Tenant::shutdown to complete, we could
attach the same tenant_id again. The shut down Tenant object might still
be around at the time of the attach.
The race is then the following:
- old object's metrics are still around
- new object uses with_label_values
- old object calls remove_label_values
The outcome is that the new object will have the metric objects (they're
an Arc internall) but the metrics won't be part of the internal registry
and hence they'll be missing in `/metrics`.
Later, when the new object gets shut down and tries to
remove_label_value, it will observe an error because
the metric was already removed by the old object.
Changes
-------
This PR moves metric removal to `shutdown()`.
An alternative design would be to multi-version the metrics using a
distinguishing label, or, to use a better metrics crate that allows
removing metrics from the registry through the locally held metric
handle instead of interacting with the (globally shared) registry.
refs https://github.com/neondatabase/neon/pull/7051
All of production is using it now as of
https://github.com/neondatabase/aws/pull/1121
The change in `flaky_tests.py` resets the flakiness detection logic.
The alternative would have been to repeat the choice of io engine in
each test name, which would junk up the various test reports too much.
---------
Co-authored-by: Alexander Bayandin <alexander@neon.tech>
Otherwise, it might happen that we never get to witness the same state
on subsequent restarts, thus the time series will show the value from a
few restarts ago.
The actual case here was that "Activating" was showing `3` while I was
doing tenant migration testing on staging. The number 3 was however from
a startup that happened some time ago which had been interrupted by
another deployment.
## Problem
Before this PR, it was possible that on-demand downloads were started
after `Timeline::shutdown()`.
For example, we have observed a walreceiver-connection-handler-initiated
on-demand download that was started after `Timeline::shutdown()`s final
`task_mgr::shutdown_tasks()` call.
The underlying issue is that `task_mgr::shutdown_tasks()` isn't sticky,
i.e., new tasks can be spawned during or after
`task_mgr::shutdown_tasks()`.
Cc: https://github.com/neondatabase/neon/issues/4175 in lieu of a more
specific issue for task_mgr. We already decided we want to get rid of it
anyways.
Original investigation:
https://neondb.slack.com/archives/C033RQ5SPDH/p1709824952465949
## Changes
- enter gate while downloading
- use timeline cancellation token for cancelling download
thereby, fixes#7054
Entering the gate might also remove recent "kept the gate from closing"
in staging.
## Problem
We reverted https://github.com/neondatabase/neon/pull/6661 a few days
ago. The change led to OOMs in
benchmarks followed by large WAL reingests.
The issue was that we removed [this
code](d04af08567/pageserver/src/tenant/timeline/walreceiver/walreceiver_connection.rs (L409-L417)).
That call may trigger a roll of the open layer due to
the keepalive messages received from the safekeeper. Removing it meant
that enforcing
of checkpoint timeout became even more lax and led to using up large
amounts of memory
for the in memory layer indices.
## Summary of changes
Piggyback on keep alive messages to enforce checkpoint timeout. This is
a hack, but it's exactly what
the current code is doing.
## Alternatives
Christhian, Joonas and myself sketched out a timer based approach
[here](https://github.com/neondatabase/neon/pull/6940). While discussing
it further, it became obvious that's also a bit of a hack and not the
desired end state. I chose not
to take that further since it's not what we ultimately want and it'll be
harder to rip out.
Right now it's unclear what the ideal system behaviour is:
* early flushing on memory pressure, or ...
* detaching tenants on memory pressure
## Problem
Closes: https://github.com/neondatabase/neon/issues/6847
Closes: https://github.com/neondatabase/neon/issues/7006
## Summary of changes
- Pageserver API calls are wrapped in timeout/retry logic: this prevents
a reconciler getting hung on a pageserver API hang, and prevents
reconcilers having to totally retry if one API call returns a retryable
error (e.g. 503).
- Add a cancellation token to `Node`, so that when we mark a node
offline we will cancel any API calls in progress to that node, and avoid
issuing any more API calls to that offline node.
- If the dirty locations of a shard are all on offline nodes, then don't
spawn a reconciler
- In re-attach, if we have no observed state object for a tenant then
construct one with conf: None (which means "unknown"). Then in
Reconciler, implement a TODO for scanning such locations before running,
so that we will avoid spuriously incrementing a generation in the case
of a node that was offline while we started (this is the case that
tripped up #7006)
- Refactoring: make Node contents private (and thereby guarantee that
updates to availability mode reliably update the cancellation token.)
- Refactoring: don't pass the whole map of nodes into Reconciler (and
thereby remove a bunch of .expect() calls)
Some of this was discovered/tested with a new failure injection test
that will come in a separate PR, once it is stable enough for CI.
## Problem
When vectored get encountered a portion of the key range that could
not be mapped to any layer in the current timeline it would incorrectly
bail out of the current timeline. This is incorrect since we may have
had layers queued for a visit in the fringe.
## Summary of changes
* Add a repro unit test
* Remove the early bail out path
* Simplify range search return value
We have a benchmark for creating a lot of branches, but it does random
things, and the branch count is not what we is the largest maximum we
aim to support. If this PR would stabilize the benchmark total duration
it means that there are some structures which are very much slower than
others. Then we should add a seed-outputting variant to help find and
reproduce such cases.
Additionally, record for the benchmark:
- shutdown duration
- startup metrics once done (on restart)
- duration of first compaction completion via debug logging
## Problem
We attempted validation for cancelled errors under the assumption that
if vectored get fails, sequential get will too.
That's not right 100% of times though because sequential get may have
the values cached and slip them through
even when shutting down.
## Summary of changes
Don't validate if either search impl failed due to tenant shutdown.
## Problem
Collection of small changes, batched together to reduce CI overhead.
## Summary of changes
- Layer download messages include size -- this is useful when watching a
pageserver hydrate its on disk cache in the log.
- Controller migrate API could put an invalid NodeId into TenantState
- Scheduling errors during tenant create could result in creating some
shards and not others.
- Consistency check could give hard-to-understand failures in tests if a
reconcile was in process: explicitly fail the check if reconciles are in
progress instead.
## Problem
- The storage controller is the source of truth for a tenant's stripe
size, but doesn't currently have a way to propagate that to compute:
we're just using the default stripe size everywhere.
Closes: https://github.com/neondatabase/neon/issues/6903
## Summary of changes
- Include stripe size in `ComputeHookNotifyRequest`
- Include stripe size in `LocationConfigResponse`
The stripe size is optional: it will only be advertised for
multi-sharded tenants. This enables the controller to defer the choice
of stripe size until we split a tenant for the first time.
## Problem
If large numbers of shards are attached to a pageserver concurrently,
for example after another node fails, it can cause excessive I/O queue
depths due to all the newly attached shards trying to calculate logical
sizes concurrently.
#6907 added the `lazy` flag to handle this.
## Summary of changes
- Use `lazy=true` from all /location_config calls in the storage
controller Reconciler.
Moves some of the (legacy) compaction code to compaction.rs. No
functional changes, just moves of code.
Before, compaction.rs was only for the new tiered compaction mechanism,
now it's for both the old and new mechanisms.
Part of #6768
