[
https://issues.apache.org/jira/browse/HDFS-7221?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14174703#comment-14174703
]
Ming Ma commented on HDFS-7221:
-------------------------------
Thanks Yongjun and Charles for investigating this.
I agree with the suggestion to increase the value for
dfs.namenode.replication.max-streams-hard-limit. Please note that
dfs.namenode.replication.max-streams normally is set to less than or equal to
dfs.namenode.replication.max-streams-hard-limit as it doesn't matter otherwise.
So as part of this fix, you can change the value for
dfs.namenode.replication.max-streams to be 16.
IMHO, per node configuration is useful if you have heterogeneous nodes in the
cluster and the scope is much more than these two properties, for example you
have other configurations such as maxXceiverCount, balancer bandwidth, etc..
Heterogeneous storages might have addressed some of the issues. Besides, it
should be easy to manage; maybe some sort of labels support in HDFS.
> TestDNFencingWithReplication fails consistently
> -----------------------------------------------
>
> Key: HDFS-7221
> URL: https://issues.apache.org/jira/browse/HDFS-7221
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: test
> Affects Versions: 2.6.0
> Reporter: Charles Lamb
> Assignee: Charles Lamb
> Priority: Minor
> Attachments: HDFS-7221.001.patch, HDFS-7221.002.patch
>
>
> TestDNFencingWithReplication consistently fails with a timeout, both in
> jenkins runs and on my local machine.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
