hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Varun Sharma (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-4960) Unnecessary .meta seeks even when skip checksum is true
Date Tue, 09 Jul 2013 21:23:49 GMT

    [ https://issues.apache.org/jira/browse/HDFS-4960?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13703810#comment-13703810

Varun Sharma commented on HDFS-4960:

Actually, I am currently failing to max out the SSD(s) and I am essentially looking out for
things that hog CPU resources or could be time sinks. Even with this patch, i fail to max
out the hardware. 

I will try to run an isolated test perhaps with just HDFS random seeks on the drives (no HBase).
I agree that we need to have a test showing that this is really better. Currently, I doubt
that it will be easy because it seems that a bunch of other paths which are hogging cpu resources.

As for SSDs, seeks do matter if you are looking max them out or utilize them well enough.
They obviously dont matter if you are looking at doing < 10K iops.
> Unnecessary .meta seeks even when skip checksum is true
> -------------------------------------------------------
>                 Key: HDFS-4960
>                 URL: https://issues.apache.org/jira/browse/HDFS-4960
>             Project: Hadoop HDFS
>          Issue Type: Bug
>    Affects Versions: 3.0.0, 2.1.0-beta
>            Reporter: Varun Sharma
>            Assignee: Varun Sharma
>         Attachments: 4960-branch2.patch, 4960-trunk.patch
> While attempting to benchmark an HBase + Hadoop 2.0 setup on SSDs, we found unnecessary
seeks into .meta files, each seek was a 7 byte read at the head of the file - this attempts
to validate the version #. Since the client is requesting no-checksum, we should not be needing
to touch the .meta file at all.
> Since the purpose of skip checksum is to also avoid the performance penalty of the extra
seek, we should not be seeking into .meta if skip checksum is true

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

View raw message