[
https://issues.apache.org/jira/browse/HADOOP-10410?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13942111#comment-13942111
]
Haohui Mai commented on HADOOP-10410:
-------------------------------------
I'm still wondering the rationales of baking the {{ioprio_*}} calls into hadoop.
>From what I get in the man page, {{ioprio_*}} calls set the I/O priority at
>the level of process / process group / user. It looks like that your
>requirement is to change the I/O priority of the HBase / HDFS daemons.
>However, these daemons are long-running processes, you can change their I/O
>priorities using a small utility.
Another issue is that changing the I/O priority of a process might need the
{{CAP_SYS_ADMIN}} capability, where most of the callees of the {{ioprio_*}
calls in hadoop do not have or do not want to have. Therefore it might make
more sense to do it through an external utility that you can run it under root.
Popping up one level, I'm wondering what {{ioprio_*}} buy you? Any real-world
performance results would be useful to guide the discussions.
> Support ioprio_set in NativeIO
> ------------------------------
>
> Key: HADOOP-10410
> URL: https://issues.apache.org/jira/browse/HADOOP-10410
> Project: Hadoop Common
> Issue Type: New Feature
> Components: native
> Affects Versions: 3.0.0, 2.4.0
> Reporter: Liang Xie
> Assignee: Liang Xie
> Attachments: HADOOP-10410.txt
>
>
> It would be better to HBase application if HDFS layer provide a fine-grained
> IO request priority. Most of modern kernel should support ioprio_set system
> call now.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
