hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Todd Lipcon (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HDFS-1148) Convert FSDataset to ReadWriteLock
Date Mon, 06 Jun 2011 05:06:47 GMT

     [ https://issues.apache.org/jira/browse/HDFS-1148?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Todd Lipcon updated HDFS-1148:
------------------------------

    Attachment: hdfs-1148-trunk.txt

Here's a patch against trunk.

When combined with HDFS-941, this yields another 15-25% or so speed-up for parts of TestParallelRead
on my machine. The CPU usage of the datanode increases from ~400% to ~750% (this is a good
thing - it's using available cores better)

> Convert FSDataset to ReadWriteLock
> ----------------------------------
>
>                 Key: HDFS-1148
>                 URL: https://issues.apache.org/jira/browse/HDFS-1148
>             Project: Hadoop HDFS
>          Issue Type: Improvement
>          Components: data-node
>            Reporter: Todd Lipcon
>            Assignee: Todd Lipcon
>         Attachments: hdfs-1148-old.txt, hdfs-1148-trunk.txt, patch-HDFS-1148-rel0.20.2.txt
>
>
> In benchmarking HDFS-941 I noticed that for the random read workload, the FSDataset lock
is highly contended. After converting it to a ReentrantReadWriteLock, I saw a ~25% improvement
on both latency and ops/second.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message