siddhantsangwan commented on PR #9362: URL: https://github.com/apache/ozone/pull/9362#issuecomment-3610694164
> I'd rather suggest graceful handling of Out of space event from DN during container creation and if it fails to use one of already opened containers. > graceful handling of Out of space I'm not sure what the client retry behaviour is. We should be having some retry mechanism to pick a different pipeline. But write performance is already degraded at that point, so we need to make better decisions at the SCM level about which pipeline is given to a client in the first place. I think we need to test the correctness and performance of `CapacityPipelineChoosePolicy` and make it the default to reduce the probability of a full pipeline being chosen. `CapacityPipelineChoosePolicy` in addition to this fix should improve the situation. -- 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]
