Network Appliance Inc. (NetApp) has stepped up its integration with database software from Oracle Corp., adding one more level of checking before data is written to a NetApp filer. The result? Significantly improved data integrity, at a price.
The product that enables this extra step is called SnapValidator, and according to NetApp, is an industry first on modular storage. IBM, EMC Corp. and Hitachi Data Systems have offered this double checksum capability on its high-end systems for a couple of years. But NetApp claims to have simplified the way it is used.
According to Phil Brotherton, senior director of marketing at NetApp, double checksum features on high-end arrays operate in "raw mode, which is for elite storage administrators only," he said. "There is no file system in raw mode, so the admin has to custom-place the data, which makes backups much more complicated." SnapValidator runs at the file level, avoiding this problem, he said.
SnapValidator is available as a separate license for Data ONTAP 7G starting at $1,000 for small filers and up to $10,000 for larger filers. Arun Taneja, founder and analyst with the Taneja Group, notes that users buying NetApp filers already assume that data reliability features are built in. "Charging extra for it would be like charging for error checking and correction (ECC), which is now part of just about all systems. Even memory has ECC built in -- no one charges for it anymore," he said.
Brotherton argued that the Oracle Hardware Assisted Resilient Data Initiative, a program designed to prevent data corruptions before they happen, "requires real work at a coding level to do the checking," but he said that over time, it's likely this feature could become free. "If we see enough adoption, it could trend toward a standard part."
By implementing Oracle's data validation algorithms inside its storage devices, SnapValidator provides Oracle on NetApp database users with an "integrity-check" by proactively checking and intercepting potential data corruption or loss. It also avoids accidental human errors, such as copying data onto inappropriate storage resources, or data corruption due to the complexity of the environment.
Without SnapValidator, users would still be able to recover their data, but would have to roll the database back to before the corruption, which takes time.