Draft fix for the OADP 4289 - to be tested before proposed upstream

[mpryc]

Draft pull request
http://quay.io/migi/velero:fix_oadp_4289

[openshift-ci]

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

[mpryc]

@sseago @shubham-pampattiwar @weshayutin if you could comment if it's proper fix.

[openshift-ci]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: mpryc

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

Needs approval from an approver in each of these files:

• DOWNSTREAM_OWNERS

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

[mpryc]

@sseago I've revisited that block of code and reworked it to be much simpler without so many if-else statements. Also checking if the error is type of conflict is unnecessary as the function RetryOnConflict won't retry on other errors.

If you could comment on current implementation.

[sseago]

The updated workflow is a lot clearer -- only one patch attempt each time the func is called. The only problems I see now are 1) handling of the obj we're adding managed fields for prior to generating patch, and 2) we shouldn't error our for NotFound errors, to replicate current behavior (where the NotFound logic was added a few releases ago to handle an edge case that a user ran into). Comments inline.

[sseago]

withManagedFields = createdObj
(I don't think we need deepcopy here, since this is the last use of createdObj)

[mpryc]

The createdObj is used later, so I wonder if we need to do some assignment so createdObj becomes updated with managed fields in a situation we get one fresh from cluster ?

[sseago]

Oh. Actually, in that case, a simpler change. Ignore the suggestion to create withManagedFields and instead, above we can set createdObj to the result of resourceClient.Get and omit the else clause completely. We'll do the deepCopy outside of the if block.

[mpryc]

Should be fixed.

[sseago]

I don't think we need the log.Errorf here, since we already logged the error inside the retry func. Also, we need to add the if block around errs.Add and return here, since we don't error out if it's a NotFound error:

	if !apierrors.IsNotFound(err) {
		errs.Add(namespace, err)
		return warnings, errs, itemExists
	}

[mpryc]

For that I think we need to also include:

	if err != nil && !apierrors.IsNotFound(err) {
		errs.Add(namespace, err)
		return warnings, errs, itemExists
	}

[mpryc]

Should be fixed now.

[sseago]

Add withoutManagedFields = createdObj.DeepCopy() here, right before calling setManagedFields.

[mpryc]

Yeah that logic is much cleaner. thanks.

[openshift-ci]

@mpryc: The following test failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
ci/prow/lint	74a4f2f	link	true	/test lint

Full PR test history. Your PR dashboard.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here.

[openshift-merge-robot]

PR needs rebase.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.