[
https://issues.apache.org/jira/browse/HDFS-10735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15457238#comment-15457238
]
Arpit Agarwal commented on HDFS-10735:
--------------------------------------
Hi [~benoyantony], the change looks fine to me. It will trigger the failover
condition in {{FailoverOnNetworkExceptionRetry}} policy.
Does checking that the wrapped exception is StandbyException work, instead of
{{re.getMessage().endsWith}}? You probably tried it already.
> Distcp using webhdfs on secure HA clusters fails with StandbyException
> ----------------------------------------------------------------------
>
> Key: HDFS-10735
> URL: https://issues.apache.org/jira/browse/HDFS-10735
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: webhdfs
> Affects Versions: 2.7.2
> Reporter: Benoy Antony
> Assignee: Benoy Antony
> Fix For: 2.8.0
>
> Attachments: HDFS-10735-001.patch, exception.txt
>
>
> A distcp running on HADOOP 2.7.2 cluster , copying from HADOOP 2.4.1 cluster
> fails.
> Both are secure clusters.
> The logs indicates that task contacts the StandBy name node on the remote
> cluster is contacted first and gets a standByException.
> Instead of retrying with the active name node on the remote cluster, the
> execution errors out throwing InvalidToken
> The exception is attached
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
