You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-dev@hadoop.apache.org by "dhruba borthakur (JIRA)" <ji...@apache.org> on 2008/08/19 10:03:45 UTC

[jira] Updated: (HADOOP-2330) Preallocate transaction log to improve namenode transaction logging performance

     [ https://issues.apache.org/jira/browse/HADOOP-2330?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

dhruba borthakur updated HADOOP-2330:
-------------------------------------

    Attachment: preallocateTransactionLog3.patch

This patch is merged with latest trunk. I have run NNThroughout benchmark on a small set of files (nrFiles =100) and it seems to show 100% improvement on "create files" test.

Hairong: I think I need both the RandomAccessFile and FileOutputStream, but both of them share the same file channel. The RandomAccessFile is used to "seek" whereas the FileOutputStream is used to write out Writables. Since both of them share the same FileChannel, it should be ok, isn't it? If you have any other way in mind, please let me know.

> Preallocate transaction log to improve namenode transaction logging performance
> -------------------------------------------------------------------------------
>
>                 Key: HADOOP-2330
>                 URL: https://issues.apache.org/jira/browse/HADOOP-2330
>             Project: Hadoop Core
>          Issue Type: Improvement
>          Components: dfs
>            Reporter: dhruba borthakur
>            Assignee: dhruba borthakur
>         Attachments: preallocateTransactionLog.patch, preallocateTransactionLog.patch, preallocateTransactionLog2.patch, preallocateTransactionLog3.patch
>
>
> In the current implementation, the transaction log is opened in "append" mode and every new transaction is written to the end of the log. This means that new blocks get allocated to the edits file frequently.
> It is worth measuring the performance improvement when big chunks of the transaction log are allocated up front. Adding new transactions do not cause frequent block allocations for the edits log.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.