hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Weiwei Yang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-12489) Ozone: OzoneRestClientException swallows exceptions which makes client hard to debug failures
Date Tue, 19 Sep 2017 15:21:00 GMT

    [ https://issues.apache.org/jira/browse/HDFS-12489?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16171884#comment-16171884
] 

Weiwei Yang commented on HDFS-12489:
------------------------------------

Hi [~linyiqun]

Thanks for the quick feedback. In my opinion, it is necessary to print stack trace when error
happens because that helps to debug. In some cases, user may not have access to server nodes
(to read logs) at all. In that case if only a short message given, it is impossible to figure
out what was the issue.

I did a quick check on other HDFS shell code, there are mostly 2 ways to handle exceptions:
1) Print out to console like this patch; 2) Log in the debug message when debug is enabled;
Since {{Shell.java}} doesn't have LOG instance yet, I simply uses #1 approach. Please let
me know if this makes sense to you.

Thanks

> Ozone: OzoneRestClientException swallows exceptions which makes client hard to debug
failures 
> ----------------------------------------------------------------------------------------------
>
>                 Key: HDFS-12489
>                 URL: https://issues.apache.org/jira/browse/HDFS-12489
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>          Components: ozone
>            Reporter: Weiwei Yang
>            Assignee: Weiwei Yang
>              Labels: ozoneMerge
>         Attachments: HDFS-12489-HDFS-7240.001.patch, HDFS-12489-HDFS-7240.002.patch
>
>
> There are multiple try-catch places swallow exceptions when transforming some other exception
to {{OzoneRestClientException}}. As a result, when client runs into such code paths, they
lose track of what was going on which makes the debug extremely difficult. See below example
> {code}
> bin/hdfs oz -listBucket  http://15oz1.fyre.ibm.com:9864/vol-0-84022 -user wwei
> Command Failed : {"httpCode":0,"shortMessage":"Read timed out","resource":null,"message":"Read
timed out","requestID":null,"hostName":null}
> {code}
> the returned message doesn't help much on debugging where and how it reads timed out.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org


Mime
View raw message