Bail out after stop timeout when stopping supervised processes #6312
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
16 tasks
twz123
force-pushed
the
bail-out-after-stop-timeout
branch
from
ed12b89 to
12a506e
Compare
Contributor
|
The PR is marked as stale since no activity has been recorded in 30 days |
twz123
force-pushed
the
bail-out-after-stop-timeout
branch
2 times, most recently
from
110ae95 to
a0ddd30
Compare
Contributor
|
This pull request has merge conflicts that need to be resolved. |
Signed-off-by: Tom Wieczorek <[email protected]>
The os.Process API is strange in that it returns an error instead of (ProcessState, error). This makes it difficult to distinguish between "regular" process errors and failures that occur while actually waiting on the process. Nevertheless, try to distinguish between these two cases to produce more accurate log messages: If the error is nil or unwraps into an *exec.ExitErr, treat it as a "regular" process error. Consider everything else an error indicating a problem with waiting. Signed-off-by: Tom Wieczorek <[email protected]>
Previously, the shutdown code looped endlessly until the child process finished, requesting graceful termination over and over again. Change this to a single request-termination -> wait -> bail-out logic. This is to ensure that k0s won't hang when the supervised processes can't be terminated for whichever reason: the code will terminate, at least after the timeout expired. Use a buffered channel for the wait result, so that the goroutine will be able to exit, even if nothing reads from the channel anymore. Introduce fine-grained error reporting to differentiate shutdown outcomes (graceful shutdown, forced kill, failure, and so on). Signed-off-by: Tom Wieczorek <[email protected]>
twz123
force-pushed
the
bail-out-after-stop-timeout
branch
from
a0ddd30 to
0827218
Compare
Member
Author
|
Oh, this PR uncovered an oversight when implementing #6429. The |
Member
Author
|
The fix for the |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
Description
Previously, the shutdown code looped endlessly until the child process finished, requesting graceful termination over and over again. Change this to a single request-termination -> wait -> bail-out logic. This is to ensure that k0s won't hang when the supervised processes can't be terminated for whichever reason: the code will terminate, at least after the timeout expired.
Use a buffered channel for the wait result, so that the goroutine will be able to exit, even if nothing reads from the channel anymore. Introduce fine-grained error reporting to differentiate shutdown outcomes (graceful shutdown, forced kill, failure, and so on).
See:
Type of change
How Has This Been Tested?
Checklist