[
https://issues.apache.org/jira/browse/HDFS-6331?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13988474#comment-13988474
]
Andrew Wang commented on HDFS-6331:
-----------------------------------
I found this comment by [~cnauroth] which explains why setACL doesn't require
the retry cache:
https://issues.apache.org/jira/browse/HDFS-5596?focusedCommentId=13846119&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13846119
However, I don't think this applies in our situation. ACLs are modified through
"blind writes" that do not depend on the current value of the ACLs, and so can
be classified as idempotent. xattr updates though are not blind writes, since
flags like CREATE and OVERWRITE depend on the currently set xattrs.
I'm happy to pick up this JIRA, but I'd appreciate if one other person could
verify my thinking first (maybe Chris?) :)
> ClientProtocol#setXattr should not be annotated idempotent
> ----------------------------------------------------------
>
> Key: HDFS-6331
> URL: https://issues.apache.org/jira/browse/HDFS-6331
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: namenode
> Affects Versions: HDFS XAttrs (HDFS-2006)
> Reporter: Andrew Wang
>
> ClientProtocol#setXAttr is annotated @Idempotent, but this is incorrect since
> subsequent retries need to throw different exceptions based on the passed
> flags (e.g. CREATE, REPLACE).
--
This message was sent by Atlassian JIRA
(v6.2#6252)
