HDDS-10482. OMRequestTestUtils.createOmKeyInfo should set key modification time

[smengcl]

What changes were proposed in this pull request?

Set mod time in OMRequestTestUtils.createOmKeyInfo to fix the regression in test util that caused test failure in a feature branch. See jira description for more backgrounds.

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-10482

How was this patch tested?

• Added assertion to prevent such regression from happening on the master branch again.

[jojochuang]

LGTM. Build is broken by a recent commit. Let's trigger a rebuild after the broken commit is reverted/fixed.

[smengcl]

I've rebased this PR to the master branch before the breaking commit. In order to run CI independent of the breakage.

[adoroszlai]

Thanks @smengcl for spotting this.

I've rebased this PR to the master branch before the breaking commit. In order to run CI independent of the breakage.

That doesn't work, PRs test patch on top of current master. Build error is now fixed on master.

[smengcl]

Thanks @smengcl for spotting this.

I've rebased this PR to the master branch before the breaking commit. In order to run CI independent of the breakage.

That doesn't work, PRs test patch on top of current master. Build error is now fixed on master.

Interesting. The trick works on my fork's branch here:

https://github.com/smengcl/hadoop-ozone/actions/runs/8181320975

I guess the PR CI just stick to the master branch when it is posted?

Anyway will rebase again.

[adoroszlai]

Interesting. The trick works on my fork's branch

Build for push (usually happens in forks, or when PR is merged) uses exactly the commit that is pushed. Build for PR works as I described.

[adoroszlai]

Thanks @smengcl for the fix.

[smengcl]

Thanks @adoroszlai for the review and comments!

[smengcl]

Thanks @jojochuang for the review as well!