zuston commented on PR #85: URL: https://github.com/apache/incubator-uniffle/pull/85#issuecomment-1196203904
> But if we have multi coordinator, we will change exclude-node-files multi times, and if we need to start shuffle server again, we need change them again. Exclude-nodes-file could be stored in HDFS, so it dont need to sync multiple coordinators. > And i think exclude is not equal to decommission I think exclude-nodes-file could be as decommission for a shuffle server. But it should support two ways of decommission 1. Normal decommission. This will make coordinator wont do assignment for this node and let it shutdown quickly. 2. Graceful decommission. This will make coordinator wont do assignment for this node and let it graceful shutdown until all apps finished. Please let me know what u think. @xianjingfeng -- 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]
