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 "Steve Loughran (JIRA)" <ji...@apache.org> on 2008/09/19 12:08:44 UTC
[jira] Commented: (HADOOP-3894) DFSClient chould log errors better,
and provide better diagnostics
[ https://issues.apache.org/jira/browse/HADOOP-3894?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12632613#action_12632613 ]
Steve Loughran commented on HADOOP-3894:
----------------------------------------
1. processDataNodeError does use the src parameter when creating an exception to forward up.
lastException = new IOException("Could not get block locations. " +
"Source file \""+src
+ "\" - Aborting...");
Adding the file that caused the error is generally useful, as it tells the caller which file is playing up. That may be immediately obvious if using the command line tools, but not when you are working with an application that is issuing lots of calls. Having that filename in test results is invaluable.
2. I managed to get the code to timeout while making changes in other parts of the system.
> DFSClient chould log errors better, and provide better diagnostics
> ------------------------------------------------------------------
>
> Key: HADOOP-3894
> URL: https://issues.apache.org/jira/browse/HADOOP-3894
> Project: Hadoop Core
> Issue Type: Improvement
> Components: dfs
> Affects Versions: 0.19.0
> Reporter: Steve Loughran
> Assignee: Steve Loughran
> Priority: Trivial
> Fix For: 0.19.0
>
> Attachments: hadoop-3894.patch, hadoop-3894.patch, hadoop-3894.patch
>
>
> In my test runs I see a stack trace from DFSClient, because it isn't logging through the log APIs in its close() method. It should use the logger, for better error reporting
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.