diegomrsantos commented on issue #4128: URL: https://github.com/apache/iggy/issues/4128#issuecomment-5751277422
@hubcio yeah, that makes sense. I added a smaller reproduction in #4130. It calls the real purge code and injects a sync failure for each offset directory. In both cases purge still succeeds and records the new generation. The control passes. It doesn't simulate power loss or check `Next` after recovering the offsets, so your proposed test would cover what's missing. I checked the retry part too. `purge_deferred` stops this replica from sending `PrepareOk`, but commits can still be applied before the next purge attempt. So we need to check that those fresh writes survive the retry. -- 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]
