advancedxy commented on PR #428: URL: https://github.com/apache/incubator-uniffle/pull/428#issuecomment-1362810583
> Maybe this is the reason that why the client write speed will be affected. It maybe lead to usedMemory release slow. @xianjingfeng ah, yeah. This is a valid concern. But we may solve flush frequency in other ways. But based on this, I will maintain the decrease per `cacheShuffleData` call. > but it's hard to avoid OOM if having too much requests come at the same time without lock. The preAllocated mechanism should already back pressure client side? So there won't be too much requests? And the coordinator should do better assignments. -- 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]
