Hi Fred
from LIP-LCG2 site, since yesterday we have having a close look at our
bdii (ii02)
which is showing very high response times (sometimes around 11 sec)
I have pointed our nodes CE/WN's to lcg-bdii.cern.ch for now.
I have a very large average load on the ii02 machine (sometimes 2) when
the bdii-update is running
we are checking what can be done
cheers
Mario
Frederic Schaer wrote:
> Hi,
>
> I resubmitted testjobs to all sites failing critical "lcg 3rd party
> replication tests" today, using another SE than lxn1183.
> It appears that the following sites still fail the test, even if there
> are fewer sites (is lxn1183 unstable ?).
>
> However, failure is often due I think to sites having a local BDII
> that might also be unstable (LCG_GFAL_INFOSYS env. variable).
> Maybe also the SEs used are unknown to the information system, because
> the BDII used excluded the site hosting the GIIS...
>
> So... SE used was : lcgse01.gridpp.rl.ac.uk
> And sites that should (IMHO) really consider looking at what's wrong
> are :
> QMUL-eScience INFN-BOLOGNA USCMS-FNAL-WC1 INFN-FRASCATI
> Hephy-Vienna NSC GSI-LCG2 INFN-NAPOLI-ATLAS
> Ru-Troitsk-INR-LCG2 LIP-LCG2
>
> Regards,
> Frederic Schaer - CIC on Duty France
|