Exagrid Systems has announced self-healing and self-protecting storage, that its claims is streets ahead of IBM.
The Exagrid Advanstor system combines a Nas primary store with disc-to-disc (D2D) backup to local or remote self-managing repositories of GRIDdisks. These Linux-based servers use about one terabyte's worth of Serial ATA drives, less expensive than the SCSI drives used in the InfiniteFiler Nas boxes.
By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
The primary store is called an InfiniteFiler, a 1U form factor Nas box. These are connected to the pool of storage in repositories. Each repository is made up of two or more GRIDdisks and is connected to the Nas filers by Gigabit Ethernet and IP. The whole system forms a storage grid in which each component is backed up by another component and data is visible across the system.
Data is stored in the system in a content-addressable form which is MD5 checksum-protected. Software in the system continually checks the validity of checksums and corrupted data is repaired by being restored from a backup held elsewhere on the grid.
Stored data on the InfiniteFiler is bled off to GRIDdisks as it gets stale according to policies, and client or application access patterns. Stubs are left behind so that it can still be accessed. There is no need for manual filestore, volume and LUN allocation activities. GRIDdisks are auto-discovered and their storage transparently added to the system's storage pool. The system scales to a large degree.
Backups are carried out automatically to discs within the system and are incremental with only changed data being backed up. There is no need for additional backup software or a tape drive-based infrastructure. Archived files are still online and recovered at disc speeds.
So what we have is a self-contained, disc-based storage utility that protects data and provides disaster recovery automatically. And that is far in advance of anything that IBM, the autonomic computing concept pioneer, has produced.
IBM has added some autonomic facilities to its storage line. The latest DS8000, for example, has a dynamically optimising and self-tuning cache to improve performance. However, it does not store data in content-addressable form with checksum protection. Nor does it automatically create copies, and so cannot self-correct corrupted data.
By inventing a new product Exagrid has advanced the autonomic storage art more than IBM can, with its need to update its existing storage line.
Patricia Murphy, vice-president of investor relations for IBM, positioned autonomic computing still as an emerging growth area under an R&D heading in a presentation in September this year. Self-healing for IBM is more a case of coping with failed hardware components, for example "fail in place", than correcting data corruption.
Advanstor is available now, priced from $62,000 (£33,380).
Chris Mellor writes for Techworld.com