INFN SC REPORT
==============

ALICE:
two activities in progress:

- DC06 production on Torino, CNAF, LNL, Catania, Bari, Cagliari, Bologna. At the end of this production, generated data will be moved to CERN.
Current status: the production is quite stable, the main issue is to keep stable our current share of CPU. We plan to go on with next week, now the production is suspended because of a problem (under investigation) at CERN.
Nothing of particular to report about italian sites involved:
ALICE doesn't have a didicated queue in Bari, but site manager are providing it;
Bologna joined the production last week after a configuration problem (solved);
There is an unstable number of job in Torino (under investigation);
Alice VOBOX in Catania was re-installed because it was unstable, now it seems better.

- FTS transfer test from CERN to INFN-T1
Current status: test in progress. Max throughput is about 32.47MB/s, with 124.4GB transfered. There are two open tickets about Castor at CNAF:  GGUS #12129 and #11541.



ATLAS:
No report.

CMS:
Investigating with CNAF storage group on Castor-2 issues, it
was clearly seen that Castor-2 services do compete quite much on
Oracle resources at CNAF, and disentangling of the resources is
now being discussed/planned. In the meantime, low transfer rates
of MTCC data (CERN->CNAF) are seen during last week, due to
Castor-2 stager unresponsiveness. On the farming side, when the
intervention on the cooling system are over, the queues have been
re-opened on friday, CMS MC production activities tried to restart
but CMS sw area was not properly mounted in the restart, it was
notified by D.Bonacorsi on friday evening, and experts intervened
immediately, all CMS MC production jobs aborted for 1.5 days until
it was fixed on Sunday. Then, success rate is again low due to
Castor-2 issues again. No SC4 transfers last week, deliberately
stopped because MC production and MTCC data transfers have higher
priority and Castor-2 apparently cannot sustain the SC4 additional
load.

LHCb:
Last week the MC production with data transfer CNAF->CERN went
on. The activity will continue this week, with no problem apart
from the down period due to the cooling issue. The best transfer
rate achieved was 10 MB/s (2-hours average). Problems are seen
in the data transfer CERN-CNAF, due to SRM getRequestStatus timing
out on put ("Error is CastorStagerInterface.c:2457 Device or
resource busy (errno=0, serrno=0)"). LHCb cannot read files back
from Castor-2 due to failure in the pre-staging from the
reconstruction jobs: the Castor-2 stager replies with a time-out
longer with the one being set-up by LHCb, which is at 30s as
suggested by GiuseppeLP during the summer (when the same problem
first arose).