Iron Mountain uses data deduplication on its repositories supporting Live Vault and Connected customers, but director of product management David Asher said data deduplication alone cannot balance the high rate of data growth.
Asher said customers have had the ability to exclude certain files from backup lists with the online backup services it offers, so files, such as JPEGs and MP3s, are skipped. "But a huge amount of material is still being captured, and too much of it doesn't get deleted," he said.
The company is developing its data classification engine based on technology of Avalere, a stealth mode startup it acquired last year. The plan is to complete a classification engine that will be used with Iron Mountain's backup SaaS offerings to create a rules-based system that can be set up by administrators or IT managers. Files can then be extracted from Iron Mountain's backup data centers and either moved to archive or deleted securely.
Iron Mountain has been doing pilot programs at customers' data centers but has not determined if it will offer the technology as a separate service or a feature upgrade to Iron Mountain's current services. The vendor offers no timeframe for bringing the product to market, although it is likely to roll out sometime next year.
Analysts said Iron Mountain hasn't yet filled in all the important details of its plan, such as exactly what role archiving will play. "I think the key [to reducing data] is to migrate the data from production to a true archive or delete at the production site, not delete it from Iron Mountain's repository," said Forrester analyst Stephanie Balaouras. "That way you get the double benefit of reducing your costs with Iron Mountain but also reclaiming production storage capacity at your own location."