[
https://issues.apache.org/jira/browse/HDFS-1480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12925206#action_12925206
]
T Meyarivan commented on HDFS-1480:
-----------------------------------
[1] verified via "dfsadmin -report" that rack id etc is being computed
correctly for all datanodes
[2] number of such replicas doesn't seem to be reducing
[3] balancer is/was running
-
[A] isGoodTarget() is broken for repl=2 ?
[B] 'hadoop fsck ...' uses slightly different logic to determine whether a
block is correctly replicated ?
--
> All replicas for a block end up in same rack
> --------------------------------------------
>
> Key: HDFS-1480
> URL: https://issues.apache.org/jira/browse/HDFS-1480
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: T Meyarivan
> Priority: Minor
>
> It appears that all replicas of a block can end up in the same rack. The
> likelihood of such replicas seems to be directly related to decommissioning
> of nodes.
> Post rolling OS upgrade (decommission 3-10% of nodes, re-install etc, add
> them back) of a running cluster, all replicas of about 0.16% of blocks ended
> up in the same rack.
> Hadoop Namenode UI etc doesn't seem to know about such incorrectly replicated
> blocks. "hadoop fsck .." does report that the blocks must be replicated on
> additional racks.
> Looking at ReplicationTargetChooser.java, following seem suspect:
> snippet-01:
> """
> int maxNodesPerRack =
> (totalNumOfReplicas-1)/clusterMap.getNumOfRacks()+2;
> """
> snippet-02:
> """
> if (counter>maxTargetPerLoc) {
> logr.debug("Node "+NodeBase.getPath(node)+
> " is not chosen because the rack has too many chosen nodes");
> return false;
> }
> """
> snippet-03:
> """
> default:
> chooseRandom(numOfReplicas, NodeBase.ROOT, excludedNodes,
> blocksize, maxNodesPerRack, results);
> }
> """
> --
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
