HDDS-5147. Intermittent test failure in TestContainerDeletionChoosingPolicy#testRandomChoosingPolicy

[elek]

What changes were proposed in this pull request?

The TestContainerDeletionChoosingPolicy#testRandomChoosingPolicy is reported to be intermittent.

Observed in this CI run: https://github.com/apache/ozone/pull/2179/checks?check_run_id=2423463857

Bundle is attached to issue.

It seems that #1885 introduced a few problems. (It was new to me, so I try to explain here the changes)

-data.incrPendingDeletionBlocks(random.nextInt(numContainers) + 1);
+data.incrPendingDeletionBlocks(20);

It's better to use fixed size of containers as it's very hard to predict the test behaviors without that.

 for (int i = 0; i < result1.size(); i++) {

That's the root cause of the problem. As we have block size based selection, the container size may be less than the requested block size.

List<ContainerBlockInfo> result1 = blockDeletingService
              .chooseContainerForBlockDeletion(50, deletionPolicy);

Here we should use block number s(50) instead of the container numbers (was 10 earlier)

for (int j = 0; j < 100; j++) {

As result1 and result2 returns with 3 containers it's better to check multiple times the randomness.

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-5147

How was this patch tested?

Unit test is passed 7000 times locally. Without the patch it fails during the first 50-100 executions.

[errose28]

Thanks for the quick fix Marton! This LGTM +1.

[lokeshj1703]

@elek Thanks for working on this! The changes look good to me. +1.

[lokeshj1703]

@elek Thanks for the contribution! @errose28 Thanks for the review! I have committed the PR to master branch.