umesh9794 commented on PR #10549: URL: https://github.com/apache/ozone/pull/10549#issuecomment-4807043992
> @umesh9794 current solution is not technically correct from production perspective. I have mentioned the reason why die to derby constraints. May be you need to think in another direction for solution where we should avoid full deletes of all records and can we bring updates to existing records and delete only those containers which gone back as Healthy. If we want to improve performance, I could foresee this is the the only safest alternative as of now. Sure @devmadhuu, I have updated the PR where Recon’s unhealthy-container sync no longer deletes all SCM state rows for every tracked container on each scan. Instead, it now diffs against what is already in the DB: insert new unhealthy rows, update rows that are still unhealthy, and delete only rows that are no longer unhealthy (for example, when a container goes back to Healthy or changes state). This avoids large bulk deletes and re-inserts, keeps Derby safe from oversized SQL predicates, and makes steady-state sync much faster when most containers stay unhealthy across cycles. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
