LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 1 Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete Tiziana Ferrari INFN CNAF CCR, Roma, Ottobre 2005

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 2 Service Challenge 3 Now in SC3 Service Phase (Sep – Dec 2005) ALICE, CMS and LHCb have all started their production ATLAS are preparing for November 1 st start Service Challenge 4: [May 2006, Sep 2006]

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 3 SC3: INFN Sites (1/2) CNAF –LCG File Catalogue (LFC) –File Transport Service (FTS) –Myproxy server and BDII –Storage: CASTOR with SRM interface Interim installations of sw components from some of the experiments (not currently available from LCG) - VObox

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 4 SC3: INFN sites (2/2) Torino (ALICE): –FTS, LFC, dCache (LCG 2.6.0) –Storage Space: 2 TBy Milano (ATLAS): –FTS, LFC, DPM –Storage space: 5.29 TBy Pisa (ATLAS/CMS): –FTS, PhEDEx, POOL file cat, PubDB, LFC, DPM –Storage space: 5 TBy available, 5 TBy expected Legnaro (CMS): –FTS, PhEDEx, Pool file cat., PubDB, DPM (1 pool, 80 Gby) –Storage space: 4 TBy Bari (ATLAS/CMS): –FTS, PhEDEx, POOL file cat., PubDB, LFC, dCache, DPM –Storage space: 5 TBy available LHCb –CNAF

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 5 INFN Tier-1: SC4 short-term plans (1/2) SC4: storage and computing resources will be shared with production Storage –Oct 2005: Data disk 50 TB (Castor front-end) WAN  Disk performance: 125 MB/s (demonstrated, SC3) –Oct 2005: Tape 200 TB (4 9940B + 6 LTO2 drives) Drives shared with production WAN  Tape performance: mean sustained ~ 50 MB/s (SC3, throughput phase, July 2005) Computing –Oct 2005: min 1200 kSI2K, max1550 kSI2K (as the farm is shared)

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 6 INFN Tier-1: SC4 short-term plans (2/2) Network –Oct 2005: 2 x 1 GEthernet links CNAF  GARR, dedicated to SC traffic to/from CERN –Future: ongoing upgrade to 10 GEthernet, CNAF   GARR, dedicated to SC Usage of policy routing at the GARR access point Type of connectivity to INFN Tier-2 under discussion Backup link Tier-1   Tier-1 (Karlsruhe) under discussion Software –Oct 2005: SRM/Castor and FTS farm middleware: LCG 2.6 –Future: dCache and StoRM under evaluation (for disk-only SRMs) Possibility to upgrade to CASTOR v2 under evaluation (end of year 2005)

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 7 What target rates for Tier-2’s? LHC Computing Grid: Technical Design Report –Access link at 1 Gb/s by the time LHC starts –Traffic 1 Tier-1   1 Tier-2 ~ 10% Traffic Tier-0   Tier-1 –Estimations of traffic Tier-1   Tier-2 represent an upper limit –Tier-2   Tier-2 replications will lower loan on the Tier-1 *WSF: with safety factor

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 8 TDR: bandwidth estimation LHC Computing Grid Technical Design Report (giugno 2005):

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 9 IN (real data Mb/s) IN (MC, Mb/s) Total IN (Mb/s) OUT (MC, Mb/s) Alice 31.9 / / / / 11.2 Atlas 32.9 / / / / 10.2 CMS 68.5 / / / / LHCb 0./ 0 0 / / 15.3 Expected rates (rough / with_safety-factor)

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 10 Necessita’ vs specifiche Tier-2 Esperimento 2006 (Mb/s) BGA/BEA in 2006 (Mb/s) BGA/BEA out Necessità totali IN / OUT (Mb/si) Bari Alice – CMS 500 / / Catania Alice 0 / 1000=118.3 / 11.2 Frascati Atlas 100 / 200=113.3 / 10.2 Legnaro Alice – CMS 450 / / / Milano Atlas 100 / 1000=113.3 / 10.2 Napoli Atlas1000 / 1000=113.3 / 10.2 Pisa CMSNon specificato / Roma1 Atlas – CMS 100 / 2500 atlas 200 / 1000 cms =318.8 / Torino Alice 70 / 1000 = / 11.2 With GigEth (Oct 2005) No Yes no Yes No Yes

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 11 The Worldwide LHC Computing Grid Level 1 Service Milestones 31 Dec 05 –Tier-0/1 high-performance network operational at CERN and 3 Tier-1s. 31 Dec 05 –750 MB/s data recording demonstration at CERN: Data generator disk tape sustaining 750 MB/s for one week using the CASTOR 2 mass storage system. Jan 06 – Feb 06 –Throughput tests SC4a 28 Feb 06 –All required software for baseline services deployed and operational at all Tier-1s and at least 20 Tier-2 sites Mar 06 –Tier-0/1 high-performance network operational at CERN and 6 Tier-1s, at least 3 via GEANT. SC4b 30 Apr 06 –Service Challenge 4 Set-up: Set-up complete and basic service demonstrated, capable of running experiment-supplied packaged test jobs, data distribution tested. 30 Apr 06 –1.0 GB/s data recording demonstration at CERN: Data generator disk tape sustaining 1.0 GB/s for one week using the CASTOR 2 mass storage system and the new tape equipment. SC4 31 May 06 Service Challenge 4: –Start of stable service phase, including all Tier-1s and 40 Tier-2 sites.

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 12 The Worldwide LHC Computing Grid Level 1 Service Milestones The service must be able to support the full computing model of each experiment, including simulation and end-user batch analysis at Tier-2 sites. Criteria for successful completion of SC4: By end of service phase (end September 2006) 1.8 Tier-1s and 20 Tier-2s must have demonstrated availability better than 90% of the levels specified the WLCG MoU [adjusted for sites that do not provide a 24 hour service] 2.Success rate of standard application test jobs greater than 90% (excluding failures due to the applications environment and non-availability of sites) 3.Performance and throughput tests complete: Performance goal for each Tier-1 is –the nominal data rate that the centre must sustain during LHC operation (200 MB/s for CNAF): –CERN-disk  network  Tier-1-tape. –Throughput test goal is to maintain for one week an average throughput of 1.6 GB/s from disk at CERN to tape at the Tier-1 sites. All Tier-1 sites must participate. 30 Sept 06 –1.6 GB/s data recording demonstration at CERN: Data generator  disk  tape sustaining 1.6 GB/s for one week using the CASTOR mass storage system. 30 Sept 06 –Initial LHC Service in operation: Capable of handling the full nominal data rate between CERN and Tier-1s. The service will be used for extended testing of the computing systems of the four experiments, for simulation and for processing of cosmic-ray data. During the following six months each site will build up to the full throughput needed for LHC operation, which is twice the nominal data rate. 24 hour operational coverage is required at all Tier-1 centres from January 2007

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 13 Milestones in brief Jan 2006: –throughput tests (rate targets not specified) May 2006: –high-speed network infrastructure operational By Sep 2006: –avg throughput of 1.6 GB/s from disk at CERN to tape at the Tier-1 sites (nominal rate for LHC operation, 200 MB/s for CNAF) Oct 2006 – Mar 2007: –avg throughput up to twice the nominal rate

LCG Service Challenge Phase 4: Piano di attività e impatto sulla infrastruttura di rete 14 Target (nominal) data rates for Tier-1 sites and CERN in SC4