hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Andrew Wang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-5016) Heartbeating thread blocks under some failure conditions leading to loss of datanodes
Date Fri, 26 Jul 2013 02:35:49 GMT

    [ https://issues.apache.org/jira/browse/HDFS-5016?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13720360#comment-13720360

Andrew Wang commented on HDFS-5016:

bq. Can you explain why it is not a good idea?

I think it'll be confusing when this gets printed in the log, since first the writer's stack
gets WARN logged, then when the exception gets caught and printed, we'll see the writer's
stack again since it's part of the exception msg and then the waiter's stack after that. Kinda
spew-y, and it makes it look like the exception was thrown from the writer since it comes
first when the exception is printed.

FWIW, we saw this triggering daily on a customer cluster. Not that common, but not that rare

bq. you want some thing like this?

Sure, that works.
> Heartbeating thread blocks under some failure conditions leading to loss of datanodes
> -------------------------------------------------------------------------------------
>                 Key: HDFS-5016
>                 URL: https://issues.apache.org/jira/browse/HDFS-5016
>             Project: Hadoop HDFS
>          Issue Type: Bug
>            Reporter: Devaraj Das
>            Assignee: Suresh Srinivas
>            Priority: Blocker
>             Fix For: 2.1.0-beta
>         Attachments: HDFS-5016.1.patch, HDFS-5016.2.patch, HDFS-5016.patch, jstack1.txt
> In the testing of some failure scenarios for HBase MTTR, we have been simulating node
failures via firewalling of nodes (where all communication ports would be firewalled except
ssh's port). We have noticed that when a (data)node is firewalled, we lose certain other datanodes
- those that were involved in some communication with the firewalled node before the latter
was firewalled. Will attach jstack output from one of the lost datanodes. The heartbeating
thread seems to be locked up.
> This jira is to track a fix for the problem.

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

View raw message