Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Added alerts for failing connectors and tasks #10315

Merged
merged 3 commits into from
Jul 11, 2024

Conversation

lhunyady
Copy link
Contributor

@lhunyady lhunyady commented Jul 8, 2024

Type of change

  • Enhancement / new feature

Description

Added alerts for failing connectors and tasks as these can not be automatically recovered and need manual intervention.

@lhunyady lhunyady marked this pull request as draft July 8, 2024 09:24
Copy link
Member

@scholzj scholzj left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for the PR.

Given users can use auto-restarting of the failed connectors / tasks, the message should be adjusted to avoid suggesting it will not recover as you don't know that.

You should also not do any changes to the examples folder as that containes only the released files.

@lhunyady
Copy link
Contributor Author

lhunyady commented Jul 8, 2024

Thanks for the PR.

Given users can use auto-restarting of the failed connectors / tasks, the message should be adjusted to avoid suggesting it will not recover as you don't know that.

You should also not do any changes to the examples folder as that containes only the released files.

Thank you for the feedback! I suppose if there is a mechanism for auto restarting the failed connectors/tasks then maybe we should wait with the alerting. I have read this article: https://strimzi.io/blog/2023/01/25/auto-restarting-connectors/ but it is not completely clear for me how much we should wait with the alert, is it ~ 30 min? how much is the backoff on the last restart cycle?

@lhunyady lhunyady marked this pull request as ready for review July 8, 2024 09:54
@scholzj
Copy link
Member

scholzj commented Jul 8, 2024

Thank you for the feedback! I suppose if there is a mechanism for auto restarting the failed connectors/tasks then maybe we should wait with the alerting. I have read this article: https://strimzi.io/blog/2023/01/25/auto-restarting-connectors/ but it is not completely clear for me how much we should wait with the alert, is it ~ 30 min? how much is the backoff on the last restart cycle?

As far as I'm concerned, failure is always a failure. So I do not think there is anything wrong with raising the alerts right away with an improved message. If you want to add some wait, I would maybe wait for the first restart only? Something like 5 minutes should maybe cover it?

@lhunyady
Copy link
Contributor Author

lhunyady commented Jul 8, 2024

As far as I'm concerned, failure is always a failure. So I do not think there is anything wrong with raising the alerts right away with an improved message. If you want to add some wait, I would maybe wait for the first restart only? Something like 5 minutes should maybe cover it?

The state will transition into "restarting" so I added for: 1m to the alerts, that should alert between the first and second restart

@scholzj
Copy link
Member

scholzj commented Jul 8, 2024

The state will transition into "restarting" so I added for: 1m to the alerts, that should alert between the first and second restart

I think the 1 minute will not make much change. After it fails in Connect, the operator will normally need 0-2 minutes to discover that it failed and restart it if needed. So 1 minute would fall right between that and give unpredictable results. So I would either set it to 2+ minutes to give the operator a chance to recover it. Or keep it as immediate.

It also looks like you opened this based on some old main branch -> I think you will need to rebase it to solve the conflcit in the CHANGELOG (the change should go under the 0.43.0 release.)

@scholzj scholzj added this to the 0.43.0 milestone Jul 8, 2024
@scholzj
Copy link
Member

scholzj commented Jul 8, 2024

Also, could you please try to fix the DCO sign-off? The instructions should be under the Details link next to the DCO status.

…omatically recovered and need manual intervention.

Signed-off-by: Laszlo I. Hunyady <laszlo.istvan.hunyady@gmail.com>
Updated the description of the alerts
Added for: 1m to the alerts to wait a loop of auto restart

Signed-off-by: Laszlo I. Hunyady <laszlo.istvan.hunyady@gmail.com>
Signed-off-by: Laszlo I. Hunyady <laszlo.istvan.hunyady@gmail.com>
Copy link
Member

@scholzj scholzj left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. Thanks.

@scholzj scholzj requested a review from a team July 8, 2024 14:34
@scholzj
Copy link
Member

scholzj commented Jul 11, 2024

Thanks for the PR.

@scholzj scholzj merged commit e5d158c into strimzi:main Jul 11, 2024
13 checks passed
mstruk pushed a commit to mstruk/strimzi-kafka-operator that referenced this pull request Jul 15, 2024
Signed-off-by: Laszlo I. Hunyady <laszlo.istvan.hunyady@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants