Increase job submission timeout and fast fail when it occurs #217
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
mwylde
requested review from
anandswaminathan,
glaksh100 and
kumare3
as code owners
|
👍 |
glaksh100
approved these changes
Oct 15, 2020
rjbordon
reviewed
Dec 4, 2020
| @@ -36,7 +37,7 @@ const httpPost = "POST" | |||
| const httpPatch = "PATCH" | |||
| const retryCount = 3 | |||
| const httpGetTimeOut = 5 * time.Second | |||
| const defaultTimeOut = 1 * time.Minute | |||
| const defaultTimeOut = 5 * time.Minute | |||
There was a problem hiding this comment.
is there a way to configure timeouts for different operations? I can see that for certain environments (minikube for instance) timeouts are prevalent.
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Outdated suggestions cannot be applied.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.
If the job submission POST request hits our 1 minute timeout (typically because the main method of the job, which needs to run to produce the job graph before job submission can complete) we can end up with two jobs submitted to the cluster.
The sequence of events looks like this:
Given that, after a timeout, we have no way to know if the job submission is still in progress on the JM, the only safe behavior in this case is to fail the deploy and rollback to the previous version.
This PR adopts that behavior, and also increases the timeout to 5 minutes which should hopefully make this much rarer in practice.