HDDS-10218. Speed up TestSstFilteringService

[raju-balpande]

What changes were proposed in this pull request?

Changes are done to improve the test run speed to 33.23 s as compare to 59.43 s earlier.

• The setup method modified to be initialize one time for all methods.
• Introduced the method orderer
• Declared one counter to track the number of snapshots been generated in all the methods to assert the count according.
• Separated the methods for creation of volume and bucket which was combine and there were repeated calls to it.
• Changed one assert statement to assert that because the .sst file from the bucket created in first method was not consistent to available in second method.

What is the link to the Apache JIRA

https://issues.apache.org/jira/browse/HDDS-10218

How was this patch tested?

Tested on local system and in CI-CD integration.

[adoroszlai]

Thanks @raju-balpande for the patch.

[adoroszlai]

Thanks @raju-balpande for updating the patch, LGTM.

[myskov]

I would avoid tests ordering.
Instead of counting snapshots across these tests, I think it will be better to get the initial snapshot number in the beginning of each test:

long initialSnapshotsNum = sstFilteringService.getSnapshotFilteredCount().get();
...
waitForSnapshotsAtLeast(filteringService, initialSnapshotsNum + 1);

This will keep these tests independent.

[hemantk-12]

I would avoid tests ordering.

+1 if it is possible.

Otherwise patch looks good to me.

[raju-balpande]

The methods are independent of ordering flow when snapshots are concerned. Independent method run also works fine. but the ordering was required because checks on key operations are dependent. Clearing all the data again will consume time and there will be hardly any improvement in time. but I am trying to see if I can fix it for keys.

[adoroszlai]

Thanks @raju-balpande for the patch, @hemantk-12, @myskov for the review.

Let's work on improving independence of test methods separately (HDDS-10346).