Antony Stubbs

Hi I suspect this is because PC doesn't know about "reset to earliest", so the underlying consumer re-downloads the records, but PC has already seen them so ignores them. If...

At least once? Are you saying PC has been skipping records?

I’m afraid this all might be been confused by issue #365 (Received duplicated records when rebalance occurs), which is fixed in release 0.5.2.1 (latest is 0.5.2.2). What is the retention...

Sounds like an idea worth exploring! Draft work will be done here: https://github.com/astubbs/parallel-consumer/compare/features/key-partition-combine...astubbs:features/multi-cluster Community let us know your thoughts!

Current workaround is to just run multiple PCs. We're thinking on what advantage supporting this inside PC might have..

Related: https://github.com/confluentinc/parallel-consumer/issues/71 Health-checks

Raised in question https://github.com/confluentinc/parallel-consumer/discussions/112

Pushed some ideas on this: https://github.com/confluentinc/parallel-consumer/compare/master...astubbs:tx-commit-failure Needs a test adding to capture some corner cases. Not a high priority as producer commit mode isn't popular, and the system will recover,...

We have a slightly friendlier error now - exceptions are caught and context given. But the original suggestions in this ticket should still be done.

Yes, that needs fixing