fix: allow stopping services in "starting" state #503
Merged
Conversation
This file contains 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
…hat quickly exist within the 1s okay delay
IronCore864
changed the title
benhoyt
changed the title
benhoyt
requested changes
Sep 26, 2024
There was a problem hiding this comment.
I'm good with the fix, thanks. However, I wonder if we can spend some time on test cases. Ideally we could have a test for each of the issues this will close:
- If a service is stopped during okayWait, it 1) returns the "stopped before the okay delay" error, and 2) is actually terminated
- If the daemon is stopped during okayWait, the service is terminated
Co-authored-by: Ben Hoyt <[email protected]>
Co-authored-by: Ben Hoyt <[email protected]>
benhoyt
reviewed
Oct 2, 2024
|
Also seems like the new daemon test is failing (sometimes). Let's ensure it's not flakey with some stress testing of the new tests. |
|
Vulnerability scanning failed, which is a known issue. Here is a workaround, shall we adopt it? I prepared a PR for this: #507. |
benhoyt
approved these changes
Oct 9, 2024
IronCore864
added a commit
that referenced
this pull request
Oct 14, 2024
…rting state (#510) In [this PR](#503), we introduced a feature to allow stopping services that are in the starting state / within the okayDelay. However, this makes [a deadlock issue in ringbuffer a real issue](#508). So we are reverting this change now and will redo it after the deadlock is resolved.
IronCore864
added a commit
that referenced
this pull request
Oct 21, 2024
Previously, we introduced [a fix to allow stopping services in the starting state](#503). Because of this fix, we discovered [another deadlock issue](#508), so we rolled it back. Now that the deadlock issue is [fixed by this PR](https://github.com/canonical/pebble/pull/511/files), we are reintroducing the fix about "allowing stopping services in the starting state". Benchmark test, manual test to reproduce the 3-way deadlock issue, and race test are all done and passed.
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.
Pebble has a 1-second okayWait time period for starting services, after which, if the service is still not exited, it's considered as "running". Previously, when services were still in the starting state within this 1-second okayWait period, they couldn't be stopped. This causes some issues in tests. And, just because services have just started doesn't mean they shouldn't be allowed to stop. This PR allows services in the starting state (within the 1s okayWait period) to be stopped, no matter if it's the daemon exits (Ctrl+C in the console) or the
pebble stopcommand is issued.Fixes #410
Fixes #502
Also adding two test cases: