Hi Santanu,
> Is there any thing has/had been changed recently in the WN middleware.
> I'm also suffering form jobs being failed without any obvious reasons
> and it stared since I upgraded all the WNs. Only I see something like
> this:
>
> LCMAPS 0: 2007-12-09.04:17:49.704038.0000004993.0000012592 :
> lcmaps_plugin_voms-plugin_run(): VOMS Signature error (failure)!
> LCMAPS 0: 2007-12-09.04:17:49.704038.0000004993.0000012592 :
> lcmaps_plugin_voms-plugin_run(): voms plugin failed
>
> looks like a problem with voms proxies for the jobs with very short run
> time. I'm just trying to figure out.
You do not have the latest version (4.7.0) of lcg-vomscerts installed:
$ globus-job-run serv03.hep.phy.cam.ac.uk /bin/rpm -q lcg-vomscerts
lcg-vomscerts-4.6.0-1
Please fix that urgently. The error message is explained here:
https://twiki.cern.ch/twiki/bin/view/LCG/TheLCGTroubleshootingGuide#VOMS_signature_error
I also note that a large fraction of the SAM tests for your site fail
with the infamous "Maradona" error. Check that all of your WNs are up
to date w.r.t. lcg-vomscerts. Other possibilities are suggested here:
http://goc.grid.sinica.edu.tw/gocwiki/Cannot_read_JobWrapper_output...
|