FrankChen021 opened a new pull request, #20292:
URL: https://github.com/apache/druid/pull/20292

   ### Description
   
   Safely publish the lazily materialized granularity interval lookup using a 
memoized supplier. Concurrent segment-allocation requests can otherwise read a 
partially populated `TreeSet` or initialize it concurrently.
   
   Observed in [PR #20291 
CI](https://github.com/apache/druid/actions/runs/34184621200/job/101930347980): 
`ITHttpInputSourceTest` rejected `2016-06-27T00:00:11.080Z` despite its 
explicit June 2016 input interval. A deterministic regression reproduces a 
missing valid interval on baseline code.
   
   ### Validation
   
   - Baseline concurrent-lookup regression failed; fixed version passed.
   - 57 focused processing tests passed; Checkstyle passed.
   - Initialization remains lazy, including empty input; existing `TreeSet` API 
unchanged.
   
   ### Release note
   
   Fix a race during granularity interval lookup initialization that could 
cause parallel ingestion to reject valid timestamps.
   
   ### Review
   
   - [x] Self-reviewed, including concurrency behavior.
   - [x] Added regression tests.
   


-- 
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