CNDB-15393: fix ShardManager#coveringRange to decrement left instead of incrementing right

[jasonstack]

What is the issue

CNDB-15393 - ShardManager#coveringRange returns a range missing left position and the ShardManager#rangeSpanned for (min, min] is not 1.

What does this PR fix and why was it fixed

Create Range with (left-1, right] to make sure both left and right are included. And make sure range span is 1 for (min, min]

[github-actions]

Checklist before you submit for review

• This PR adheres to the Definition of Done
• Make sure there is a PR in the CNDB project updating the Converged Cassandra version
• Use NoSpamLogger for log lines that may appear frequently in the logs
• Verify test results on Butler
• Test coverage for new/modified code is > 80%
• Proper code formatting
• Proper title for each commit staring with the project-issue number, like CNDB-1234
• Each commit has a meaningful description
• Each commit is not very long and contains related changes
• Renames, moves and reformatting are in distinct commits
• All new files should contain the DataStax copyright header instead of the Apache License one

[blambov]

Nit: I would let wraparound take care of the min->max transition as in nextValidToken.

Note: Any decorated key would have a non-minimum token, so wraparound here could not happen for sstable start keys.

[jasonstack]

fixed in 26932c4

[blambov]

Hmm. MINIMUM is not a real ring position, but it makes better sense for what we are trying to do with this method as an adjustment for a ZERO token for an SSTable's first key. There is some risk that wrapping around will cause trouble in some function that is only written to work with non-wraparound ranges: these normally accept minimum as the upper bound, but not maximum on the lower side.

I would change this to go from ZERO to MINIMUM and either not touch MINIMUM or wrap it around. I know this means MAXIMUM.nextValidToken().prevValidToken() != MAXIMUM, but that's something I'd more easily accept (especially if we add a comment).

[jasonstack]

good point. fixed in 26932c4

[jasonstack]

UnifiedCompactionStrategyTest.testGetBucketsDifferentWsUniqueArena unit test failed because it uses min token as DecoractedKey for SSTable#first key (DecoratedKey first = new BufferDecoratedKey(partitioner.getMinimumToken(), ByteBuffer.allocate(0))) and DecoratedKey#isMinimal is always false. Thus ShardManager#coveringRange retrun (max, max] which is a wrap around range.

I have pushed e71ac0f to check if token is minimal instead of checking if partition-position is minimal. @blambov wdyt?

[sonarqubecloud]

Quality Gate passed

Issues
3 New issues
0 Accepted issues

Measures
0 Security Hotspots
100.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[cassci-bot]

❌ Build ds-cassandra-pr-gate/PR-2003 rejected by Butler

---

6 regressions found
See build details here

---

Found 6 new test failures

Test	Explanation	Runs	Upstream
o.a.c.db.compaction.UnifiedCompactionStrategyTest.testGetBucketsDifferentWsUniqueArena[useDiskBoundaries false] (compression)	REGRESSION	🔵🔴	0 / 5
o.a.c.db.compaction.UnifiedCompactionStrategyTest.testGetBucketsDifferentWsUniqueArena[useDiskBoundaries true] (compression)	REGRESSION	🔵🔴	0 / 5
o.a.c.db.compaction.UnifiedCompactionStrategyTest.testGetBucketsSameWUniqueArena[useDiskBoundaries false] (compression)	REGRESSION	🔵🔴	0 / 5
o.a.c.db.compaction.UnifiedCompactionStrategyTest.testGetBucketsSameWUniqueArena[useDiskBoundaries true] (compression)	REGRESSION	🔵🔴	0 / 5
o.a.c.index.sai.cql.VectorUpdateDeleteTest.testTTLOverwriteHasCorrectOnDiskRowCount[db] (compression)	REGRESSION	🔴🔵	0 / 5
o.a.c.net.ProxyHandlerConnectionsTest.suddenDisconnect (compression)	REGRESSION	🔵🔴	0 / 5

---

No known test failures found