VEP-25: Incremental backup with CBT - FF Extension Request

[Adi Aloni]

Hello,

We are requesting a one-week extension following the 1.8 FF to merge the PRs required to complete VEP 25:
https://github.com/kubevirt/kubevirt/pull/16426
https://github.com/kubevirt/kubevirt/pull/16883
https://github.com/kubevirt/kubevirt/pull/16528
https://github.com/kubevirt/kubevirt/pull/16399
https://github.com/kubevirt/kubevirt/pull/16877
https://github.com/kubevirt/kubevirt/pull/16817
https://github.com/kubevirt/kubevirt/pull/16886

Additionally, we will submit one final critical PR by the end of this week. We aim to have the existing PRs approved by 24.02.2026, and the final PR approved by 02.03.2026.

[Michael Henriksen]

I know this has been a big effort and I support an extension

--
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/4c9b24a8-0d6d-4b82-bb3a-4f3385cbb16an%40googlegroups.com.

[Itamar Holder]

I have to admit this situation is concerning.

We're talking about:

• 7 PRs that did not merge yet. Additional one that was not even submitted yet.
• Out of these, only 2 PRs have review comments on them.
• We are less than a week before the code-freeze.
• Although we have asked people to ask for reviews in time, this was not raised in any of the public meetings (community meeting, VEP discussion, etc).

The fact that a red flag was not raised about this in time is unfortunate.

I wasn't very involved in the VEP to be honest, but I think that rushing these in carries a high risk.

To view this discussion visit https://groups.google.com/d/msgid/kubevirt-dev/CAF0u-ngr0butCaDrj1185mpFrcFXcaFZC4ffrOkUx2Z1EQNFvw%40mail.gmail.com.

[Luboslav Pivarc]

Hi Adi,

I tend to agree with Itamar here, it seems concerning to see such volume of PRs one week before code freeze. Even worse, these touch a lot of components and it's not nicely isolated as one would hope :) 

What further worries me is that some PRs opened last year!

Therefore my question is do you have a sufficient reviewers pool for your effort? Did you raise this issue in a community meeting or within a VEP meeting? (Or within SIG for that matter?)

Let's postpone the decision to the 25th, if by the time the majority of code gets merged and a reasonable amount of work will be left to merge I tend to lean forward granting the exception. I ask reviewers to not lower their expectations on the code just to make this in :) 

Also please update the VEP tracker, it's left without attention for too many days.

Best of luck,

-Lubo

To view this discussion visit https://groups.google.com/d/msgid/kubevirt-dev/CALpNySwcfxiTxFbpWJpM7FdsdrtihH70L8B7oiXRjmKyiTooWg%40mail.gmail.com.

[Vladik Romanovsky]

Hi Adi,

Thank you for this effort. This VEP is genuinely one of the most requested storage features and will be a big win for the project.

However, looking at the PRs, considering the volume and size makes me very concerned about landing this in 1.8
Several XXL PRs were opened very recently; others have few reviews, missing tests.
The whole chain looks very complex and the additional critical PR hasn't even opened yet.

Rushing this risks destabilizing the release.
We've seen this pattern before, where late large features almost always destabilize the release.

I'm surprised that no attention was raised or review requests for this work earlier in the cycle.

We have weekly community calls, dedicated VEP discussion calls, (other calls too), slack, exactly for coordinating large, complex features early.

Posting most of the implementation in the final 2–3 weeks before FF makes it extremely difficult to thoroughly review these kinds of changes.

Are we sure that all the posted PRs are required for the Alpha implementation?

Thanks,

Vladik

To view this discussion visit https://groups.google.com/d/msgid/kubevirt-dev/CAF0u-ngr0butCaDrj1185mpFrcFXcaFZC4ffrOkUx2Z1EQNFvw%40mail.gmail.com.

[Adi Aloni]

Hi all,

First of all thank you for raising your concerns. Indeed, my grooming of this VEP and engagement with the community (including the VEP meetings) were lacking, and I will definitely improve that with future VEPs. That said, the majority of the PRs I listed in the original email have already been merged or approved and are pending merge. The final piece is the pull mode PR[1] which was submitted earlier today. We consider pull mode imperative to the overall feature and need it to reach alpha in 1.8.

Given the progress of the prerequisite PRs, would it be possible to reconsider our request for a 1 week extension? 

[1] https://github.com/kubevirt/kubevirt/pull/16930

Thanks,
Adi

--

Adi Aloni

Senior Software Engineer, CNV Storage Platform

Red Hat

[Shelly Kagan]

Hi all,

As a co-owner of this VEP, I am joining the request for a one-week extension to merge Adi's final PR.

This PR is a vital component of the feature and has generated significant interest from the community. As this is an Alpha feature, it remains protected behind a feature gate. Furthermore, the changes are largely contained to a new backup output method specific to the incremental backup functionality, the rest of which has already been merged.

Given the substantial effort invested in this VEP and the value it provides to users, I believe a seven-day extension is worthy.

Thank you,

Shelly Kagan

To view this discussion visit https://groups.google.com/d/msgid/kubevirt-dev/CAEN50_u70Z0fBnqQCsMvRc65u-QyQXQaimoseBLCd%2Bbg1guyig%40mail.gmail.com.

--

Shelly Kagan

Senior Software Engineer

Red Hat

[Michael Henriksen]

Hi All,

Very good progress has been made on this VEP in the last week. All previously mentioned PRs have been reviewed and approved.

Yesterday, Adi submitted the "critical" final PR for pull mode backup that brings everything together: https://github.com/kubevirt/kubevirt/pull/16930

I've done an initial review and Adi has already responded to my comments. If we maintain this momentum, I think we can reasonably expect this PR to be merged within the extension period.

Thank you for your consideration.

Mike

[Vladik Romanovsky]

Thank you for the update! 

Looks like the future of this work is in good hands!

+1 From me

Thank you!

Vladik