anew opened a new pull request, #56434:
URL: https://github.com/apache/spark/pull/56434

   ### What changes were proposed in this pull request?
   
   All AutoCDC E2E test suites share a single static
   SharedTablesInMemoryRowLevelOperationTableCatalog.sharedTables map and 
previously used the same namespace "ns1". When SBT runs suites in parallel 
within the same JVM (parallelExecution = true), one suite's afterEach calling 
reset() (which clears the entire map) destroys tables that another concurrent 
suite is actively using, causing flaky failures where stale data survives full 
refreshes or auxiliary state persists.
   
   Fix:
   - Give each suite a unique namespace derived from its class name so table 
Identifier keys never collide across suites.
   - Replace the global reset() in afterEach with a targeted clearNamespace() 
that removes only this suite's entries from the shared map.
   - Compute the ExpectedContext stop position dynamically in 
AutoCdcScd1SchemaEvolutionSuite so it adapts to the longer namespace.
   
   ### Why are the changes needed?
   To fix a flaky test that often fails the CI
   
   ### Does this PR introduce _any_ user-facing change?
   No
   
   ### How was this patch tested?
   It fixes a test
   
   ### Was this patch authored or co-authored using generative AI tooling?
   Generated-by: Opus 4.6


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to