[ https://issues.apache.org/jira/browse/HDFS-2290?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13091668#comment-13091668 ]
Konstantin Shvachko commented on HDFS-2290: ------------------------------------------- *The correct behavior* should be that NN replicates the block using one of the good replica, and then deletes the corrupt replica. So there are 2 steps: # replicating a good replica # deleting the corrupt replica when the block is fully replicated Replication is handled by ReplicationMonitor, which replicates blocks from {{neededReplications}} list. It seems that while processing corrupt replicas we somehow miss to put them into {{neededReplications}}. > Block with corrupt replica is not getting replicated > ---------------------------------------------------- > > Key: HDFS-2290 > URL: https://issues.apache.org/jira/browse/HDFS-2290 > Project: Hadoop HDFS > Issue Type: Bug > Components: name-node > Affects Versions: 0.22.0 > Reporter: Konstantin Shvachko > Fix For: 0.22.0 > > > A block has one replica marked as corrupt and two good ones. countNodes() > correctly detects that there are only 2 live replicas, and fsck reports the > block as under-replicated. But ReplicationMonitor never schedules replication > of good replicas. -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira