I too was facing the same error. But I was able to run fsck successfully from the namenode. Hope this helps.
MalikHusain wrote: > > Were you able to resolve this? I am running into a similar issue. I am > currently evaluating a 3 node cluster, When I do "hadoop fsck /" on the > namenode or one of the datanodes it successfully shows ths status as > healthy. However, on the third machine (datanode) the fsck command stopped > working after a reboot. It throws an Exception with a Connection refused. > Can you please let me know how I can resolve it. Everything else on the > datanode works except for the fsck command. > > Error on the third node > hadoop fsck / > Exception in thread "main" java.net.ConnectException: Connection refused > -- View this message in context: http://old.nabble.com/getting-HDFS-to-rack-aware-mode-tp19980091p29724233.html Sent from the Hadoop core-user mailing list archive at Nabble.com.
