Mailing-List: contact core-dev-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: core-dev@hadoop.apache.org
Message-ID: <1173251056.1236284396235.JavaMail.jira@brutus>
Date: Thu, 5 Mar 2009 12:19:56 -0800 (PST)
From: "Raghu Angadi (JIRA)" <jira@apache.org>
To: core-dev@hadoop.apache.org
Subject: [jira] Commented: (HADOOP-4865) Trigger block scans for datanode
In-Reply-To: <613388103.1229283824233.JavaMail.jira@brutus>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HADOOP-4865?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12679335#action_12679335 ] 

Raghu Angadi commented on HADOOP-4865:
--------------------------------------


looks better now. I think you need to make sure these 'manual scan' blocks are scanned without a throttle, otherwise they will be scanned pretty slowly. That is what I would want DN to do if I am asking them scan asap.

nit: it is better rename 'manuallyScanList' to 'manualScanList'?

bq. > I think it is no good to run scan process in another thread, since it's possible that the regular scan process will scan same block with the manually scan progress.

may be.. if you are fine with latencies of the order of minutes for start of scan of once such a block is queued. 

> Trigger block scans for datanode
> --------------------------------
>
>                 Key: HADOOP-4865
>                 URL: https://issues.apache.org/jira/browse/HADOOP-4865
>             Project: Hadoop Core
>          Issue Type: New Feature
>          Components: dfs
>            Reporter: Brian Bockelman
>         Attachments: manual_block_scan.patch, manual_fsck_scan.patch
>
>
> Provide a mechanism to trigger block scans in a datanode upon request.  Support interfaces for commands sent by the namenode and through the HTTP interface.

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