Add Tolerations to Build and BuildRun objects

[dorzel]

Changes

Fixes #1636

Submitter Checklist

• Includes tests if functionality changed/was added
• Includes docs if changes are user-facing
• Set a kind label on this PR
• Release notes block has been filled in, or marked NONE

See the contributor guide
for details on coding conventions, github and prow interactions, and the code review process.

Release Notes

Add Tolerations to Build and BuildRun objects

[dorzel]

Bit stumped on the failing integration tests - after several revisions I can't find why the taint effect isn't getting set. It also looks like the e2e tests are running out of disk space.

[openshift-ci]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign heavywombat for approval. For more information see the Code Review Process.

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

Needs approval from an approver in each of these files:

• OWNERS

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

[SaschaSchwarze0]

Can you explain this. There is a BuildRun which cannot start because the tolerations do not match any node. Is not the result that the Pod is stuck in Pending, eventually the TaskRun and BuildRun time out. Why would there be a failure coming from the Git source step ?

[dorzel]

Ah, this was an oversight on my part. I would expect the Pending/timeout you mentioned instead. I'll change this.

[SaschaSchwarze0]

Okay, you may want to create the buildrun with a shorter timeout. I guess it would otherwise wait the default ten minutes.

[dorzel]

An update - still looking into why TaintEffect isn't getting set in the integration tests. Integration tests are at least all failing in the same way.
e2e tests are all over the place, with unrelated tests failing and 2/4 of them nearly passing. I'm wondering if some of these failures are due to the new three-node configuration and disk space issues.

[dorzel]

@SaschaSchwarze0 @adambkaplan

Would I be able to get another look at this? I'm having an issue with the validation logic for the taint effect:
https://github.com/shipwright-io/build/pull/1711/files#diff-991ca77e26cb9e18ce83c2dc437ab9dd148e6c9263b933267733b1b3c0b45c7dR49-R58

In the case where the effect isn't specified in the yaml, as in this test:
https://github.com/shipwright-io/build/pull/1711/files#diff-1cc6608c14aa73c7ad391f5f3d05b858cef1e22b084e778ce651e5b5e663b46aR594

It should get explicitly set on the TaskRun object via the validation logic, but that doesn't happen. Confused, as everything else for the validation seems to work.
Wondering if I'm missing something obvious here or if I need to revert to setting it in taskrun.go

[SaschaSchwarze0]

Would I be able to get another look at this? I'm having an issue with the validation logic for the taint effect:
https://github.com/shipwright-io/build/pull/1711/files#diff-991ca77e26cb9e18ce83c2dc437ab9dd148e6c9263b933267733b1b3c0b45c7dR49-R58

Sry for the late reply @dorzel.

The validation logic should imo NOT mutate anything on the object. It should only validate. If empty is a good value for the effect, then that's all the validation should do: not complain anything.

The logic that creates the toleration on the TaskRun should then again be aware that empty is a valid value for the effect and translate that into TaintEffectNoSchedule.

[dorzel]

Ok, reverted to setting the TaintEffect outside of validation and that has fixed the integration tests (one unrelated flaky test). Looks like we're just waiting on #1786 to get an actual signal from e2e tests.