Tre2878 commented on PR #5855: URL: https://github.com/apache/hadoop/pull/5855#issuecomment-1661545892
> > Two of the cases will come into this logic > > > > 1. The namenode is restarting while receiving FBRS from all datanodes and is in safe mode > > 2. When the namenode is in secure mode for some reason while it has been running for a long time > > In the first case, if the datanode has a failed disk, the datanode will send the FBR for the normal disk and the namenode will handle it normally > > In the second case, blockReportCount == 0 will always be false if no new disks are added to the datanode > > So I recommend keeping the code as it is and not using blockReportCount == 0 > > @Tre2878 If a disk is failed, its' state will be set to `FAILED` and it will be removed from `storageMap`. So the check of `blockReportCount` will not involve failed storages, and `blockReportCount == 0` is not always be false. Thus, I think my plan can work here. What's your opinion? You're talking about the first case I mentioned, which is fine, The problem is the second case, where the namenode is not restarted, but the namenode is in safe mode for some reason, and this logic will also be executed, and since this is not the first FBR, blockReportCount will always be greater than 0 -- 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: common-issues-unsubscr...@hadoop.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: common-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: common-issues-h...@hadoop.apache.org
