[
https://issues.apache.org/jira/browse/HADOOP-13283?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15340119#comment-15340119
]
Mingliang Liu commented on HADOOP-13283:
----------------------------------------
Thanks, [[email protected]]. I also believe reset is preferred for production
code use case. When I discuss this with [~hitesh], I think from Tez's point of
view, the reset operation is good enough by now (correct me if I'm wrong
[~hitesh]). However, I do believe snapshot is a nice-to-have feature,
especially for testing/debugging cases. As it's non-trivial (to me) to take
care of the memory overhead and consistency, and basically it needs a different
implementation, I think we can address this in separate JIRAs later.
> Support reset operation for new global storage statistics and per FS storage
> stats
> ----------------------------------------------------------------------------------
>
> Key: HADOOP-13283
> URL: https://issues.apache.org/jira/browse/HADOOP-13283
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: fs
> Reporter: Mingliang Liu
> Assignee: Mingliang Liu
> Fix For: 2.8.0
>
>
> Applications may reuse the file system object across jobs and its storage
> statistics should be reset. Specially the {{FileSystem.Statistics}} supports
> reset and [HADOOP-13032] needs to keep that use case valid.
> This jira is for supporting reset operations for storage statistics.
> Thanks [~hitesh] for reporting this.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
