[ https://issues.apache.org/jira/browse/HBASE-16716?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15570594#comment-15570594 ]
Pankaj Kumar commented on HBASE-16716: -------------------------------------- Test case failure is not relevant, locally it is successful. Findbug warning is also not relevant, {noformat} Return value of java.util.concurrent.CountDownLatch.await(long, TimeUnit) ignored in org.apache.hadoop.hbase.regionserver.HRegionServer.initializeZooKeeper() {noformat} > OfflineMetaRepair leaves empty directory inside /hbase/WALs which remains > forever > --------------------------------------------------------------------------------- > > Key: HBASE-16716 > URL: https://issues.apache.org/jira/browse/HBASE-16716 > Project: HBase > Issue Type: Bug > Components: hbck > Affects Versions: 2.0.0 > Reporter: Pankaj Kumar > Assignee: Pankaj Kumar > Priority: Minor > Fix For: 2.0.0 > > Attachments: HBASE-16716-V2.patch, HBASE-16716-branch-1.patch, > HBASE-16716.patch > > > OfflineMetaRepair rebuild Meta table, while creating meta region it creates > it's own WAL (inside /hbase/WALs/hbck-meta-recovery-<randomNumber>) which wll > be closed and archived after rebuilding Meta. > {noformat} > hbase org.apache.hadoop.hbase.util.hbck.OfflineMetaRepair > >> /hbase/WALs/hbck-meta-recovery-<randomNumber> > {noformat} > It doesn't clear the empty dir, empty directory should be removed after > success. -- This message was sent by Atlassian JIRA (v6.3.4#6332)