[
https://issues.apache.org/jira/browse/HAMA-585?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13285741#comment-13285741
]
Suraj Menon commented on HAMA-585:
----------------------------------
Thanks Chiahung. This is a big task that you removed from my to-do list!
I think this should be a part of HAMA-505. I would be merging this to my
current changes for HAMA-534.
I just perused through your patch. Can the function, moveToBlackList(String
host) in BSPMaster call the task recovery logic or should I keep a Thread
watching blackList? I don't know how the call to blacklist node is going to be
made. The recovery logic would run if any tasks were found running on the
blacklisted GroomServer.
> Increase capability for master to be notified when a groom server fails.
> ------------------------------------------------------------------------
>
> Key: HAMA-585
> URL: https://issues.apache.org/jira/browse/HAMA-585
> Project: Hama
> Issue Type: Sub-task
> Components: bsp core
> Reporter: ChiaHung Lin
> Assignee: ChiaHung Lin
> Attachments: HAMA-585.patch
>
>
> Enhance the bsp master function by allowing master to be notified when
> detecting a groom server fail.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
