HDDS-4199. Fix failed UT: TestOMAllocateBlockRequest#testValidateAndUpdateCache

[runzhiwang]

What changes were proposed in this pull request?

What's the problem ?

What's the reason ?

we can not make sure omKeyInfo.getCreationTime() different from omKeyInfo.getModificationTime()

1. the creationTime was set in OmKeyInfo at TestOMRequestUtils.addKeyToTable -> OmKeyInfo#createOmKeyInfo

2. the modificationTime was set in KeyArgs at doPreExecute(createAllocateBlockRequest()) -> OMRequest#preExecute.
   And the KeyArgs.modificationTime was set in OmKeyInfo at omAllocateBlockRequest.validateAndUpdateCache -> openKeyInfo.setModificationTime.

3. Between doPreExecute(createAllocateBlockRequest()) and TestOMRequestUtils.addKeyToTable, there is no time consuming code, so we can not make sure the millisecond time is different.

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-4199

How was this patch tested?

no need new test.

[amaliujia]

Thanks for the investigation!

I am not sure if this is a good practice: will it help to insert a wait/sleep/a random operation after doPreExecute(createAllocateBlockRequest()); to create the opportunity to differentiate creation time and modification?time

[runzhiwang]

@bharatviswa504 Could you help review this patch ? Thank you very much.

[bharatviswa504]

Instead of completely removing, can we check greater than or equal to?

[amaliujia]

In fact, greater than or equal to is a good idea!

[runzhiwang]

@bharatviswa504 @amaliujia Thanks for review. I have updated the patch. Besides, I change the order of doPreExecute(createAllocateBlockRequest()) and TestOMRequestUtils.addKeyToTable, to make sure creationTime <= modificationTime, otherwise creationTime >= modificationTime looks weird.

[runzhiwang]

@bharatviswa504 CI passed. Could you help merge this patch ? Thank you very much.

[adoroszlai]

Thanks @runzhiwang for the fix.

[adoroszlai]

Thanks @runzhiwang for fixing this, @amaliujia and @bharatviswa504 for the review.