hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Yongjun Zhang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-9181) Better handling of exceptions thrown during upgrade shutdown
Date Tue, 06 Oct 2015 15:13:27 GMT

    [ https://issues.apache.org/jira/browse/HDFS-9181?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14945159#comment-14945159

Yongjun Zhang commented on HDFS-9181:

Hi Guys,

Since HDFS-7533 tries to catch NPE:
When datanode is told to shutdown via the dfsadmin command during rolling upgrade, it may
not shutdown. This is because not all writers have responder running, but sendOOB() tries
anyway. This causes NPE and the shutdown thread dies, halting the shutdown after only shutting
down DataXceiverServer.

I'd think that we just change the Throwable to Exception as Daniel suggested, and issue a
warning message with the exception's information. For any other thing thrown from the sendOOB
step, the behavior will be just like before HDFS-7533.


> Better handling of exceptions thrown during upgrade shutdown
> ------------------------------------------------------------
>                 Key: HDFS-9181
>                 URL: https://issues.apache.org/jira/browse/HDFS-9181
>             Project: Hadoop HDFS
>          Issue Type: Improvement
>            Reporter: Wei-Chiu Chuang
>            Assignee: Wei-Chiu Chuang
>            Priority: Minor
> Previously in HDFS-7533, a bug was fixed by suppressing exceptions during upgrade shutdown.
It may be appropriate as a temporary fix, but it would be better if the exception is handled
in some other ways.
> One way to handle it is by emitting a warning message. There could exist other ways to
handle it. This jira is created to discuss how to handle this case better.
> Thanks to [~templedf] for bringing this up.

This message was sent by Atlassian JIRA

View raw message