fix: move pubsub publishing out of database transactions to avoid conn exhaustion #17648
+155
−105
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…n starvation Signed-off-by: Danny Kopping <dannykopping@gmail.com>
6efe39f to
ba2f90a
Compare
Signed-off-by: Danny Kopping <dannykopping@gmail.com>
Signed-off-by: Danny Kopping <dannykopping@gmail.com>
dannykopping
commented
May 2, 2025
| clock: clock, | ||
| registerer: registerer, | ||
| done: make(chan struct{}, 1), | ||
| provisionNotifyCh: make(chan database.ProvisionerJob, 10), |
There was a problem hiding this comment.
Thumb-suck; I want to protect against temporary network/database blips but also don't want to accumulate too many messages.
johnstcn
reviewed
May 2, 2025
…ot affect behaviour Signed-off-by: Danny Kopping <dannykopping@gmail.com>
Signed-off-by: Danny Kopping <dannykopping@gmail.com>
johnstcn
reviewed
May 2, 2025
| func (*brokenPublisher) Publish(event string, _ []byte) error { | ||
| // I'm explicitly _not_ checking for EventJobPosted (coderd/database/provisionerjobs/provisionerjobs.go) since that | ||
| // requires too much knowledge of the underlying implementation. | ||
| return xerrors.Errorf("refusing to publish %q", event) |
There was a problem hiding this comment.
Should we fail slowly instead?
e.g. <-time.After(testutil.IntervalFast)
|
@dannykopping is it possible to use another approach, and allow calling func (p *PGPubsub) PublishWithinTx(event string, message []byte, tx database.Store) error {
tx.ExecContext(context.Background(), `select pg_notify(`+pq.QuoteLiteral(event)+`, $1)`, message)
}Implementations which doesn't support txs may implement it: func (p *InMemPubSub) PublishWithinTx(event string, message []byte, tx database.Store) error {
_ = tx // ignored
return p.PublishTx(event, message)
}But it will require some changes in Because we impose restrictions on ourselves, every time we need to use Publish within tx - we have to come up with workaround like in this PR. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Database transactions hold onto connections, and
pubsub.Publishtries to acquire a connection of its own. If the latter is called within a transaction, this can lead to connection exhaustion.I plan two follow-ups to this PR:
https://github.com/coder/coder/blob/main/cli/server.go#L2360-L2376
We will then be able to write tests showing how connection exhaustion occurs.
pubsub.Publishfrom being called within a transaction.