Assegnazione risorse Stato INFN CNAF 22 12 2004

Assegnazione risorse Stato INFN CNAF, 22 -12 -2004

Risorse attuali CPU 320 biproc. 0. 8 -3 GHz 350 biproc. 3 GHz (+70 server) 1300 KSi 2 K +55 Babar +48 CDF +30 LHCb Disco FC, IDE, SCSI, NAS 200 TB (~32/40 TB free) +35 TB (? ) Tapes Stk L 180 Stk 5500 18 TB 6 LTO-2 1200 cassette=240 TB 2 9940 B 680 cassette=136 TB Rete 29 Switch da rack 2 core Switch 46 FE UTP + 2 GE FO 96 GE FO + 128 GE FO e 4 x 10 GE

Statistiche PBS 1/3 -16/12

Disk Status (1) HW Tot Assigned Unass. Procom nas 2 9. 7 TB Virgo (9. 7 free 0) virgo-gateway-1 nfs 0 Procom nas 3 4. 2 TB Virgo (4. 2 free 1) virgo-gateway-1 nfs 0 STK Blade. Store 15 x 0. 92 TB 5 x 1. 8 TB ~25 TB (22, 8 TB) CMS (8 x 0, 92 TB + 5 x 1, 8 TB = ~18 TB (16, 36 TB) free ~3 TB) diskservcms-2 gridftp CMS (2 x 0, 92 TB) castor staging disksrv-2 AMS (0, 92 TB free 0, 2) diskserv-ams-1 nfs ATLAS (2 x 0, 92 TB) castor staging disksrv-1 2 x 0. 92 TB infortrend-1 ~3 TB ALICE castor staging disksrv-4 0 infortrend-2 ~3 TB ALICE diskserv-alice-1 rfio 0 infortrend-3 ~3 TB LHCB castor staging disksrv-3 0 infortrend-4 ~3 TB 0 ~3 TB IBM Fast-1 24 x 1, 9 TB ~48 TB (45, 6 TB) ATLAS (4 x 1, 9 TB = ~8 TB free ~5 TB ) autofs nfs ATLAS (2 x 1, 9 TB = ~4 TB) castor staging disksrv-rfio-1 LHCB (3 x 1, 9 TB = ~6 TB free ~2 TB ) autofs nfs LHCB (8 x 1, 9 TB = ~16 TB) castor staging disksrv-rfio-2 7 x 1, 9 TB IBM Fast-2 24 x 1, 9 TB ~48 TB (45, 6 TB) CMS (4 x 1, 9 TB = ~8 TB free ~6, 5 TB) gpfs diskserv-lhc-2 gridftp CDF (12 x 1, 9 TB = ~24 TB free ~17, 5 TB) gpfs rfio diskserv-cdf-2 8 x 1, 9 TB IBM Fast-3 24 x 1, 9 TB ~48 TB (45, 6 TB) BABAR (17 x 1, 9 TB= ~34 TB) diskserv-san-9/10 xroot ALICE (1 x 1, 9 TB= ~2 TB) castor staging disksrv-san-11 VIRGO (2 x 1, 9 TB= ~4 TB free 1 TB) diskserv-san-12 nfs 4 x 1, 9 TB

Disk Status (2) HW Tot Assigned Unass. Dell 660 f ~7 TB REMOVED FROM HEP PRODUCTION (need Windows for dell support) 7 TB Raidtec ~2 TB ALL castor staging disksrv-5 (generic stager) 0 Fibrereach ~2, 5 TB CDF (no monitoring on raid) diskserv-cdf-1 nfs 0 Axus Brownie ~2 TB REMOVED FROM PRODUCTION (raid controller problem possible data loss!) 0 nas 1 ~2 TB MAGIC (1 x 1. 1 TB) nfs 1 x 1. 1 TB nas 4 ~3 TB TO BE INSTALLED. . . 3 TB In progress. . . • Upgrade of the 150 TB IBM Fast storage ~24 TB not ready until end January 2005 (fast-4) • Tender for 200 TB RAW ~150 TB probably not before 2 nd half of 2005 • Upgrade of the bladestore and IBM Fast storage (see above) is possible if needed (max 15/20 TB) in 20/30 days period • Also small upgrade of the infortrends (max 3/6 TB each) is possible if needed in 20/30 days period

Disk status (exp. view) EXP Tot Assigned N. Stager/Diskserver/NAS ALICE ~8 TB 1, 5 TB rfio. 6, 5 TB staging area 3 (diskserv-san-11, diskserv-alice-1, disksrv-4) ATLAS ~14 TB 8 TB nfs. 6 TB staging area 3 (diskserv-san-1, diskserv-san-3, disksrv-1) BABAR ~34 TB xroot 2 (diskserv-san-9, diskserv-san-10, disksrv-10) CDF ~26, 5 TB 24 TB rfio, 2, 5 TB nfs 2 (diskserv-cdf-1, diskserv-cdf-2) CMS ~24 TB 22 TB gridftp. 2 TB staging area 3 (diskserv-cms-2, diskserv-lcg-2, disksrv-2) LHCB ~25 TB 6 TB nfs. 19 TB staging area 3 (diskserv-san-2, diskserv-san-4, disksrv-4) VIRGO ~18 TB nfs 3 (nas 2, nas 3, diskserv-san-12) MAGIC, AMS have 1 TB with dedicated diskserver/nas

Tape Status EXP Tot Used N. Stager/castor diskserver ALICE (also ALICE TOF) ~6 TB 30 x 200 GB LTO-2 3 (diskserv-san-11, diskserv-alice-1, disksrv-4) ATLAS ~6 TB 30 x 200 GB LTO-2 2 (diskserv-san-3, disksrv-1) CMS ~6 TB (50% wasted) 60 x 200 GB LTO-2 (many problems in write and random read) 1 (disksrv-2) LHCB ~16 80 x 200 GB LTO-2 2 (disksrv-3, diskserv-san-4) ARCHIVE (qld, lvd) ~1 TB 5 x 200 GB LTO-2 1 (disksrv-5) + soon 2 TB 1 diskserver N. 6 LTO-2 drives with 1200 x 200 GB LTO-2 tapes => 240 TB (used only 15%!) N. 2 9940 B drives with 680 x 200 GB 9940 B tapes => 136 TB (free) Our silos will be upgraded to 5500 slots (2000 LTO form factor + 3500 9940 B form factor) Upgrade with other 4 x 9940 B drives and 2500 x 200 GB tapes (500 TB) not before 2 nd half of 2005

Punti di discussione • Uso CPU, al momento, non critico (anzi…) – Occorre uniformare accesso per ottimizzare uso • Piu’ critico lo spazio disco – ~ 32 TB disponibili (+ ~ 26 TB con estensione 20% ultima gara) • Gia’ in arrivo richieste… – CASTOR alternativa valida per compensare • Buffer temporanei e/o distribuiti su dischi WNs – Importante avere previsione di uso “vera” • Pianificazione a breve/medio termine con esperimenti