Backport of "Add Prometheus metric for guest OS panic events"

[Michal Skrivanek]

Hi,
I’d like to check about back porting https://github.com/kubevirt/kubevirt/pull/17836 to 1.8. This is very useful fairly small change to bubble up metrics of crashed guests. Specifically for Windows BSODs this allows detecting those and creating alerts when the happens.
@jean-edouard as the approver of that PR on main, do you have any concerns?

Thank you,
michal

[Jed Lejosne]

Thanks for your email Michal! This is not a straight bug fix so, as stated in https://github.com/kubevirt/kubevirt/blob/main/docs/release-branch-backporting.md, it may "require some debate". That said, I'd argue it falls under "Logging and debug changes aimed at improving supportability of a stable release", and seeing that BSODs are both extremely critical and really hard to debug, I think this backport makes a lot of sense.

Please open a PR and cc me. In the meantime, if anyone has an objection, please speak up!

Thank you,

Jed

[Luboslav Pivarc]

Hi Michal,

A concern was raised on the PR, I would like to see this being addressed even before considering a backport.

-Lubo

--
You received this message because you are subscribed to the Google Groups "kubevirt-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kubevirt-dev...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/kubevirt-dev/CACDz6wDNGGvYUavVk_gyQHxbKC8oc4ipbXHbtp71FagJuaz43w%40mail.gmail.com.

[Itamar Holder]

On Fri, 26 Jun 2026 at 16:59, 'Jed Lejosne' via kubevirt-dev <kubevi...@googlegroups.com> wrote:

Thanks for your email Michal! This is not a straight bug fix so, as stated in https://github.com/kubevirt/kubevirt/blob/main/docs/release-branch-backporting.md, it may "require some debate". That said, I'd argue it falls under "Logging and debug changes aimed at improving supportability of a stable release", and seeing that BSODs are both extremely critical and really hard to debug, I think this backport makes a lot of sense.

Please open a PR and cc me. In the meantime, if anyone has an objection, please speak up!

To be perfectly honest, I had a different idea of the policy's meaning when reading it, especially this section:

Bug Fix Only: The backport must be a bug fix and the bug fix must be first merged into the main branch. The only exception is when a bug only exists in a stable branch and does not exist in the main branch.

 

The part you quoted was under the "bug fix definition" section. I tend to think that the mentioned PR is more of a feature than a bug, even if it helps identify bugs by alerting of Windows BSODs.

That being said, I agree there's some degree of ambiguity here, and the policy itself mentions this is in purpose:

There is purposefully some ambiguity here. These are meant to be read as guidelines to help reviewers make judgment calls

Therefore, if others feel this should be eligible for backport I won't resist.

I'd be happy to hear more thoughts on this.

BR,

Itamar.

Thank you,

Jed

On Fri, Jun 26, 2026 at 9:49 AM Michal Skrivanek <mskr...@redhat.com> wrote:

Hi,
I’d like to check about back porting https://github.com/kubevirt/kubevirt/pull/17836 to 1.8. This is very useful fairly small change to bubble up metrics of crashed guests. Specifically for Windows BSODs this allows detecting those and creating alerts when the happens.
@jean-edouard as the approver of that PR on main, do you have any concerns?

Thank you,
michal

[Michal Skrivanek]

On 26. 6. 2026, at 18:24, Luboslav Pivarc <lpi...@redhat.com> wrote:

Hi Michal,

A concern was raised on the PR, I would like to see this being addressed even before considering a backport.

Hi Lubo,

Seems it’s not an issue really. https://github.com/kubevirt/kubevirt/pull/17836#issuecomment-4831091014

The follow-up test improvement would also be nice, I guess, but that should be all

Thanks,

Michal