INFN Tier-1 site report
=======================
New FTS monitor web page available at http://argus.cnaf.infn.it/fts/index-FTS.php

20061023 ALICE weekly report
No report

20061023 ATLAS weekly report
No report

20061023 CMS weekly report
by Daniele Bonacorsi

Last week was CSA06 "week-3", and focus was on 1) data transfers + 2) first skim job submission. --- 
1) Data transfers of primary-reconstructed RECO from the T0 to the T1 went smoothly, and tape migration on Castor-2 continued, with no backlog. On Friday, D.Bonacorsi asked for an increase of rate to INFN T1 and the ad-hoc Castor set-up allowed to reach >80 MB/s stable for several hours (with peak at >170 MB/s) (Oct 20: plot 1) and again >80 MB/s stable (Oct 21: plot 2), then the T1 went out of data to be delivered from CERN.
Transfers to the T2 went on all week, with different efficiencies in different channels, though. On Sunday, a name-server related overload caused by a monitoring implemented with an uncorrect policy blocked the system for some hours, and it was restored on monday morning. --- 
2) Skim jobs started to flow since tuesday, and - after a WMS problem preventing the CE matchmaking to work proficiently - it went smooth. It still has a peak-like shape on T1 farm, and has not a stable load over the CPUs, with not more then some hundreds of jobs/day. It will increase this week ("week-4"), when data transfers T0-T1-T2s continues, but with more focus on the skimming jobs.
Plots are available in http://littleblue.cnaf.infn.it/twiki/bin/view/Sandbox/RocScReport200610#ScRep20061023

20061023 LHCB weekly report
by Angelo Carbone

Last week report:
MCProduction: Ongoing without problem, but  now the CNAF(T1s) is used mainly for reconstruction and Analysis, hence  the rate of MC simulation jobs was very low.
Data Transfer: No activities due to an error  on  LHCb software application.
Reconstruction job: The Reconstruction is running fine, only few jobs crashed due to temporary down of Castor2.

Plans for the current week:
The MC production will continue but with a very low rate. The data transfer will continue with the usual bandwidth (10MB/s every two our).
The reconstruction will continue increasing the rate of running jobs (increasing the read rate trough Castor2).