Abrevity Inc., based in San Jose, Calif. and boasting just 12 employees, claims most products for ILM don't help customers benefit from tiered storage, because they rely on file-system metadata that is too broad to provide the granular classification ILM requires. It's a well documented problem (see Veritas's Future Is Scrutinized). Abrevity's alternative is a package named FileBase, which it says is far more granular than current products.
On the face of it, the claim is puzzling because Abrevity's FileBase seems to work like several other ILM packages: It extracts information about files from Windows servers and uses that to create and implement file policies, such as which files should be moved to cheaper long-term storage.
Abrevity claims its secret is being able to recognize and parse directory path information in CIFS-based file systems. According to Abrevity cofounder Joel Harrison, formerly of Quantum Corp. (NYSE: DSS), using file systems to create metadata for ILM is limited by what the file system can report -- usually, a handful of criteria, such as file extension. FileBase goes way beyond this, Harrison says, not only using the file-system criteria but sorting through the different elements of a file name to find consistencies with other names, and matching that data to what's found in multiple directory paths.
Initially, Abrevity is selling FileBase as a service with an option to buy the software. At least one customer has signed on for the software, and several more are in the pipeline, Harrison says.