Naman-Katiyar commented on PR #41328:
URL: https://github.com/apache/beam/pull/41328#issuecomment-6890489031

   Thank you for the review — both points are addressed in this rework:
   
   - Wrong sink: I have reverted the PubsubSink change entirely, and moved the 
fix to `PubsubIO.Write.PubsubBoundedWriter`. It now passes `getIdAttribute()` 
to the Pub/Sub client in `startBundle` and supplies a record id in 
`processElement`.
   - Deterministic id: you are right that `UUID.nameUUIDFromBytes` would give 
identical payloads identical ids, risking silent deduplication. The bounded 
writer now generates a fresh per-element `UUID.randomUUID()` whenever a message 
lacks the configured id attribute, and keeps any id the message already carries.
   
   The tests have moved with the fix: the PubsubSinkTest additions are removed, 
and PubsubIOTest now has a capturing-client test for the batch writer covering 
fresh, distinct ids for two identical-payload elements and preservation of an 
existing id.
   
   @damccorm — could you please take another look when you have a chance?


-- 
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]

Reply via email to