High-energy physics experiments are currently recording large amounts of dataand in a few years will be recording prodigious quantities of data. New methodsmust be developed to handle this data and make analysis at universitiespossible. Grid Computing is one method; however, the data must be cached at thevarious Grid nodes. We examine some storage techniques that exploit recentdevelopments in commodity hardware. Disk arrays using RAID level 5 (RAID-5)include both parity and striping. The striping improves access speed. Theparity protects data in the event of a single disk failure, but not in the caseof multiple disk failures. We report on tests of dual-processor Linux Software RAID-5 arrays andHardware RAID-5 arrays using a 12-disk 3ware controller, in conjunction with250 and 300 GB disks, for use in offline high-energy physics data analysis. Theprice of IDE disks is now less than $1/GB. These RAID-5 disk arrays can bescaled to sizes affordable to small institutions and used when fast randomaccess at low cost is important.
展开▼
