[
https://issues.apache.org/jira/browse/MAPREDUCE-2999?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13108992#comment-13108992
]
Thomas Graves commented on MAPREDUCE-2999:
------------------------------------------
So it doesn't appear that registering "/" or the redirect I have implemented
now actually fully work because they both block the base log servlet from
serving /logs.
I think whats happening is that if you try to go to /logs it goes to the logs
servlet (AdminAuthorizedServlet.class), then it goes to the filters. Since
guice has the "/" registered guice hijacks it and tries to dispatch it on its
own and the logs servlet never gets it back.
Now I changed the pathSpec on the /logs to /*
(logContext.addServlet(AdminAuthorizedServlet.class, "/*")). Then the guice
filter doesn't hijack it and logs is served up. I don't quite understand why
this works as I would have thought the pathSpec being served was still "/" and
guice filter would have grabbed it. Any servlet experts out there that can
explain that?
Another option might be to not have guice register "/" and have an index.html
that redirects it similar to how jobtracker and namenode do now.
> hadoop.http.filter.initializers not working properly on yarn UI
> ---------------------------------------------------------------
>
> Key: MAPREDUCE-2999
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-2999
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2
> Affects Versions: 0.23.0
> Reporter: Thomas Graves
> Assignee: Thomas Graves
> Fix For: 0.23.0
>
> Attachments: MAPREDUCE-2999.patch, MAPREDUCE-2999.patch,
> MAPREDUCE-2999.patch
>
>
> Currently httpserver only has *.html", "*.jsp as user facing urls when you
> add a filter. For the new web framework in yarn, the pages no longer have the
> *.html or *.jsp and thus they are not properly being filtered.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
