There was a fair bit of red on GRidmap yesterday, but less in other
countries than UK. Does anyone know how to look at the particular test
that was failing across many sites at once?
NIKHEF (well, Jeff) was complaining recently about the large amount of
stuff that ATLAS were publishing into the BDII.
John
-----Original Message-----
From: Testbed Support for GridPP member institutes
[mailto:[log in to unmask]] On Behalf Of Matt Hodges
Sent: 13 June 2010 08:17
To: [log in to unmask]
Subject: Re: lcg-bdii.gridpp.ac.uk problem?
>>>>> catalin condurache writes:
> We believe the problem is rectified now. It appeared that CERN-PROD
> was pumping too much information into the BDII, much more than what
> was expected to be processed by the top BDII cache at RAL T1. Our
> BDII experts increased some cache limits and now
> # ldapsearch -x -h lcg-bdii.gridpp.ac.uk -p 2170 -b o=grid
> glueseuniqueid=samdpm001.cern.ch | grep ID:
> GlueSEUniqueID: samdpm001.cern.ch
> We expect to see the SAM OPS vo tests for the CE passing OK soon, and
> also the the UK wide problems to disappear.
To provide a bit more detail, the CERN-PROD ldif was rejected by the
info provider script (glite-info-provider-ldap) because it exceeded 5
MB. The limit can be increased by calling with the following arg:
-s Maximum file size in megabytes for a single source.
and to alleviate yesterday's problems, I doubled the limit to 10 MB.
I'll get someone to follow up and see if this is configurable in YAIM,
or is different for the gLite 3.2 release (which we held back on rolling
out due to stability issues).
Matt
|