HDDS-13776. Fail DirectoryPurge requests if previous snapshot ID validation fails.

[SaketaChalamchala]

What changes were proposed in this pull request?

If a snapshot is created for a bucket while DirectoryDeletingService is processing deleted directories from the same bucket then, the DirectoryPurge requests submitted by DirectoryDeletingService should not execute and should be processed when snapshot is deep cleaned.

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-13776

How was this patch tested?

[Copilot]

Pull Request Overview

This PR addresses a concurrency issue between directory deletion and snapshot creation by failing DirectoryPurge requests when snapshot ID validation fails. When a snapshot is created for a bucket while DirectoryDeletingService processes deleted directories from the same bucket, the DirectoryPurge requests should fail and be reprocessed later during snapshot deep cleaning.

• Changes the previous snapshot ID validation from void to boolean return type
• Adds conditional logic to fail the request when validation returns false

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

The error message 'Snapshot validation failed' is too generic and doesn't provide enough context for debugging. Consider adding more specific information about what aspect of validation failed or why the request should be retried.

Suggested change

	return new OMDirectoriesPurgeResponseWithFSO(createErrorOMResponse(omResponse,
	new OMException("Snapshot validation failed", OMException.ResultCodes.INVALID_REQUEST)));
	}
	String actualPrevSnapshotId = (fromSnapshotInfo != null && fromSnapshotInfo.getPreviousSnapshot() != null)
	? fromSnapshotInfo.getPreviousSnapshot().toString() : "null";
	String errorMsg = String.format(
	"Snapshot validation failed for fromSnapshot='%s': expected previousSnapshotId=%s, actual previousSnapshotId=%s",
	fromSnapshot,
	expectedPreviousSnapshotId != null ? expectedPreviousSnapshotId.toString() : "null",
	actualPrevSnapshotId
	);
	return new OMDirectoriesPurgeResponseWithFSO(createErrorOMResponse(omResponse,
	new OMException(errorMsg, OMException.ResultCodes.INVALID_REQUEST)));

[jojochuang]

Good catch @SaketaChalamchala
This is caused by HDDS-12982

So we have two options
(1) work on this RP, or
(2) revert HDDS-12982

Regardless, it looks like we need a test to validate this scenario, otherwise we could have caught the bug.

[smengcl]

Good catch @SaketaChalamchala This is caused by HDDS-12982

So we have two options (1) work on this RP, or (2) revert HDDS-12982

Regardless, it looks like we need a test to validate this scenario, otherwise we could have caught the bug.

Thanks @SaketaChalamchala for the fix.

+1 that we'd better add a test case.

[jojochuang]

I'll go ahead and merge it. If we want to add a test we can add later.

[ivandika3]

@jojochuang @SaketaChalamchala This seems to break build in master.

[adoroszlai]

Reverted due to compile error, which happened because the corresponding change in SnapshotUtils was reverted in #9162.

Please do not merge with outdated CI run. Approving pending CI run after 2 weeks still tests the change against master at the time of PR creation.

[swamirishi]

@SaketaChalamchala @jojochuang we don't need this since HDDS-13799 is already reverts the initial change. Maybe a PR with test case would be better