[
https://issues.apache.org/jira/browse/HDFS-5017?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13715801#comment-13715801
]
Jing Zhao commented on HDFS-5017:
---------------------------------
The patch looks good to me. Some thoughts after discussing with [~sureshms]:
1. Maybe we do not need to refresh a cache entry's access time and position in
the priority queue when the entry is accessed, since the expected timeout on
the client side is based on the time the first request is sent.
2. It would be better if we can add an upper limit for the size of the GSet.
This can guarantee NN continues to function even when a large amount of retry
requests come within a short period of time.
> Extends LightWeightGSet to support eviction of expired elements
> ---------------------------------------------------------------
>
> Key: HDFS-5017
> URL: https://issues.apache.org/jira/browse/HDFS-5017
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: namenode
> Reporter: Tsz Wo (Nicholas), SZE
> Assignee: Tsz Wo (Nicholas), SZE
> Attachments: h5017_20130721.patch, h5017_20130722b.patch,
> h5017_20130722.patch
>
>
> We extends LightWeightGSet to support eviction of expired elements so that
> namenode could use it as the RetryCache.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
