taklwu commented on pull request #2113: URL: https://github.com/apache/hbase/pull/2113#issuecomment-662866824
yup, it should be only the case of ALL WALs (user table WAL and MasterRegion WALs) has been flushed but removed, it should have no data loss. You're right that the new set of region servers start on new environment that they are identified as 'Unknown server'. although it's not common on long running cluster, this is very common on a one-time cluster in a dynamic environment but StoreFiles and Tables has been persisted. HBCKSCP and `scheduleRecoveries` were the manual recovery, but our point were comparing to branch-1.4+ that didn't have this issues where `dead` and `unknown` servers were expired and regions can be assigned (and if I found the history correctly, that logic were removed from [processofflineServersWithOnlineRegions -> processOfflineRegions in HBASE-20708](https://github.com/apache/hbase/commit/3e33aecea2da1071d7336584ed29448c233fe3be#diff-322aa95013b832abcc3fa2a396a70748L1310-L1357). ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected]
