[
https://issues.apache.org/jira/browse/NUTCH-1202?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13149694#comment-13149694
]
Ferdy Galema commented on NUTCH-1202:
-------------------------------------
Do you mean initialization as in Fetcher.configure() for Nutch and
FetcherMapper.setup(Context) for Nutchgora?
This would break the "relative" part. I'm referring to the following comments
in Fetcher:
// set the actual time for the timelimit relative
// to the beginning of the whole job and not of a specific task
// otherwise it keeps trying again if a task fails
Or do you mean actual MapReduce task initialization sort of? Implying that this
is only done once (even when tasks are failing) this would keep the semantics
as described above.
> Fetcher timebomb kills long waiting fetch jobs
> ----------------------------------------------
>
> Key: NUTCH-1202
> URL: https://issues.apache.org/jira/browse/NUTCH-1202
> Project: Nutch
> Issue Type: Bug
> Components: fetcher
> Reporter: Markus Jelsma
> Fix For: 1.5
>
>
> The timebomb feature kills of mappers of jobs that have been waiting too long
> in the job queue. The timebomb feature should start at mapper initialization
> instead, not in job init.
> Thoughts?
--
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
