ivanandika98 commented on PR #4371: URL: https://github.com/apache/ozone/pull/4371#issuecomment-1463592795
Hi @prashantpogde @hemantk-12 @GeorgeJahad I see that there is already an effort to introduce [incremental checkpoint](https://github.com/apache/ozone/pull/3980) in the OM snapshot process. Our cluster is currently encountering issue in which a slow OM follower has to download a large OM metadata due to the leaders' log being purged. This merge request seeks to circumvent this issue by disabling `raft.server.log.purge.upto.snapshot.index` so that leader will only purge the log once the followers have replicated the log. Could you take a look? However this has a risk in which the OM leader disk space could be filled quickly if the follower is down / very slow. Therefore I think the long-term solution would be to integrate incremental checkpoint. May I know what is the progress on the feature? We are interested in integrating it in our cluster. -- 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]
