[ https://issues.apache.org/jira/browse/HBASE-21806?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16756740#comment-16756740 ]
Sergey Shelukhin edited comment on HBASE-21806 at 1/31/19 12:46 AM: -------------------------------------------------------------------- Yeah based on a test run in our cluster I think the value for the new parameter could be around 3-5sec. The timeout could also be shorter although I see it will cause RS to crash when the timeout is hit during memstore flush WAL write, so I'm not sure if making it very short is a good idea. I will keep the new setting it at the conservative 10sec for now. was (Author: sershe): Yeah based on a test run in our cluster I think the value for the new parameter could be around 3-5sec. The timeout could also be shorter although I see it will cause RS to crash when the timeout it hit during memstore flush WAL write, so I'm not sure if making it very short is a good idea. I will keep the new setting it at the conservative 10sec for now. > add an option to roll WAL on very slow syncs > -------------------------------------------- > > Key: HBASE-21806 > URL: https://issues.apache.org/jira/browse/HBASE-21806 > Project: HBase > Issue Type: Bug > Reporter: Sergey Shelukhin > Assignee: Sergey Shelukhin > Priority: Major > Attachments: HBASE-21806.patch > > > In large heterogeneous clusters sometimes a slow datanode can cause WAL syncs > to be very slow. In this case, before the bad datanode recovers, or is > discovered and repaired, it would be helpful to roll WAL on a very slow sync > to get a new pipeline. > Otherwise the slow WAL will impact write latency for a long time (slow writes > result in less writes result in the WAL not being rolled for longer) -- This message was sent by Atlassian JIRA (v7.6.3#76005)