KEP-4443: Add more granular Job failure reason for PodFailurePolicy

[danielvegamyhre]

• One-line PR description: Configurable Job failure reason for PodFailurePolicy

• Issue link: Add more granular failure reason for Job PodFailurePolicy #4443

• Other comments:

[danielvegamyhre]

@soltysh for sig-apps lead review
tagging @ahg-g @alculquicondor @kannon92 for review as well

[danielvegamyhre]

@deads2k would you be able to do an API review for this? Here we are proposing adding an optional Reason field to the PodFailurePolicyRule for the Job API, similar to the Reason field I proposed here in the Job success policy KEP.

[kannon92]

You need a prod-readiness file for alpha I believe.
keps/sig-apps/prod-readiness/4443.yaml

[danielvegamyhre]

@wojtek-t would you mind doing a PRR review for this KEP?

[soltysh]

There are still some missing bits that needs to be filled in before we can accept this KEP.

[soltysh]

PodFailurePolicy reason was rejected quite some time ago, as potentially overbloating condition's reason field. Index or the short name, which would be validated to fit in the condition's reason was preferred.

[soltysh]

Still holds.

[soltysh]

Clarifying the default is the last bit to get this approved.

[soltysh]

The above although readable to human is hard to parse. The idea is to be able to use the reason being in your example PodFailurePolicy_0 and parse it in higher level controller.

[soltysh]

I'm of the opinion we should not default, leave it empty, which is acceptable in which case the resulting Reason will contain the index. Iow. no defaulting happening, and that should be explicitly mentioned here.

[danielvegamyhre]

Clarifying the default is the last bit to get this approved.

Updated the defaulting section to specify there will be no defaulting, and if Name is unset the Job controller will use the rule index in the condition reason suffix instead of the name.

[soltysh]

One point from Filip, but that's not blocking as long as it gets implemented.

/lgtm
/approve
from sig-apps pov

/hold
for prr review

/assign @jpbetz

[soltysh]

/lgtm

[jpbetz]

/approve
For PRR

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: danielvegamyhre, jpbetz, soltysh

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• keps/prod-readiness/OWNERS [jpbetz,soltysh]
• keps/sig-apps/OWNERS [soltysh]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[ahg-g]

/hold cancel

Thanks a lot all for the great feedback, I think we got to a very good conclusion here! kudos to the community!