somandal commented on PR #16341: URL: https://github.com/apache/pinot/pull/16341#issuecomment-3085909486
> @somandal The ZK access part makes sense to me, thanks for clarifying. > > Agree with the removal of `forceDownload` and the removal of the check that blocks rebalances at first sight on peer-tables with `downtime=true` or `minAvailableReplicas=0`. > > I think the one you mentioned in option 1 to disable the data loss protection is a fair point, in case that they don't care about data loss. Though the philosophy is not totally aligned with `bestEffort`, where `bestEffort` is saying "I want to continue rebalancing even though data may have lost", versus "I'm okay with data loss" here. IMHO we might want a new flag here. > > Plus if we observed some performance issue, this flag can be utilized to make trade-off. Done, replaced the flag. @J-HowHuang can you help with looking into adding some tests for this new flag (i.e. try to emulate a case where with downtime we fail rebalance due to ZK metadata but we pass it when the new flag is enabled) if possible? I'm running out of time to add 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: commits-unsubscr...@pinot.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org
