You are viewing a plain text version of this content. The canonical link for it is here.
Posted to hdfs-issues@hadoop.apache.org by "Kihwal Lee (JIRA)" <ji...@apache.org> on 2017/08/01 14:27:02 UTC

[jira] [Commented] (HDFS-12232) ConfiguredFailoverProxyProvider caches DNS results indefinitely

    [ https://issues.apache.org/jira/browse/HDFS-12232?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16108988#comment-16108988 ] 

Kihwal Lee commented on HDFS-12232:
-----------------------------------

I believe it can be worked around by enabling ipc client level retry, which is disabled by default in HA setup.  This allows the internal addr change logic to re-resolve and connect to the new server.  The fundamental fix is more involved as storing and handling of {{ConnectionID}} need to be changed.

> ConfiguredFailoverProxyProvider caches DNS results indefinitely
> ---------------------------------------------------------------
>
>                 Key: HDFS-12232
>                 URL: https://issues.apache.org/jira/browse/HDFS-12232
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: ha, hdfs
>            Reporter: Haohui Mai
>
> {{ConfiguredFailoverProxyProvider}} resolves the IP addresses and stores them in the internal state. As a result when migrating the standby NN to a different machine, it requires restarting all HDFS clients.
> A better approach is to follow Java's policy of caching DNS names: http://javaeesupportpatterns.blogspot.com/2011/03/java-dns-cache-reference-guide.html



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org