rohankumardubey commented on PR #3957: URL: https://github.com/apache/iggy/pull/3957#issuecomment-5550942090
> one open question: the delivery failures section (README lines 77-82) makes the 5-nack stop the operator contract. with the default 10s poll and a table that keeps producing rows, any iggy outage longer than five polls stops the source for good and needs a manual restart, since the empty-poll ack that used to reset the counter no longer happens. was exempting send-failure nacks from the sdk counter ruled out, or is manual restart the intended contract? @hubcio It wasn’t ruled out. I chose the documented manual-restart behavior from the earlier two options to keep the change scoped, but I agree that stopping after a short Iggy outage is harsh. Since send failures currently reach the SDK as an undifferentiated NACK, exempting them would require extending the batch-result contract. Would you prefer that change in this PR? -- 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]
