Over the last few months, vendors like EMC Corp., Archiware and TimeSpring Software Corp. have released products with continuous data protection (CDP), and storage experts say this capability is offers a considerable advantage over periodic backups, which may not restore the most recent data.
“It’s huge for people in Interac and that sort of thing, where if you lose a transaction, if you’re down for a couple of seconds, you could be losing thousands and thousands of dollars,” said Laura Hansen, research analyst with London, Ont.-based Info-Tech Research Group.
Info-Tech frequently gets queries from its clients on CDP, Hansen said, but added there’s some confusion in the marketplace.
She said a true CDP product does not simply take snapshots of data and applications at various points.
“Snapshots are point in time copies that you would run on a schedule, so you would have a snapshot every hour, every four hours, that sort of thing,” Hansen said. “CDP is more like snapshots on steroids, so as soon as data is changed it’s going to be backed up on CDP. If you had to restore data, you’d have only minutes or seconds of lost data whereas with snapshots, it could be an hour or four hours of lost data.”
And some companies can’t afford to lose anything they’ve saved, whether it’s their e-mail or their notes, said Patty Then, senior product marketing manager for storage management at Islandia, N.Y.-based CA Inc.
“Just about everyone needs some level of CDP for a variety of different reasons,” Then said. “Everyone needs access to their data, even what we would think is the most insignificant, because it becomes part of doing business.”
CA announced in February XOsoft High Availability r12 and XOsoft Replicationr12, both of which include a CDP repository designed to make it easier to get recently-saved data in the event of a failure.
The XOsoft brand was retained from a company of the same name bought by CA.
Another company that acquired a CDP maker was EMC Corp. of Hopkinton, Mass., which two years ago acquired Israeli vendor Kashya Inc. EMC has since integrated Kashya’s technology into its storage offerings, and recently unveiled version 3 of RecoverPoint. The product now includes network-based replication and lets administrators decide if they want to replicate data to a local device or a remote storage site.
“Previously, for any one dataset, I had an either/or decision in how I want to protect my data — locally or remotely,” said Rick Walsworth, EMC’s director of product marketing, who was originally with Kashya.
Walsworth said the remote and local protection gives administrators more flexibility. “Let’s say in the case of a logical corruption, I have a transactional database, a corruption event took place,” he said. “I can recover that using the local copy. On top of that, if I have a flood or something catastrophic that took place in my primary data center, I have a copy of that data at the remote site that I can use to recover from.”
He added database administration is a major driver of the CDP market, and it’s even more attractive as executives get a better appreciation of their e-mail archives.
“They started with the Oracle implementation, they put it in place, it’s working, then the Exchange (database administrator) looks over and says, ‘How can I provide the same level of protection?'”
With RecoverPoint, EMC aims to give administrators the ability to define the importance of different applications.
That Oracle DBA may need a 15-second protection window — in other words, my remote site will always be within 15 seconds of my primary site,” Walsworth said. “The Exchange environment may not be as mission-critical so you can define another service level.”
Users need to figure out how long they want their data to be available before it’s logged as a snapshot, Hansen said. She added companies implementing CDP need to have a “solid bandwidth” connection so they can actually move the data they’re saving, and they need to figure out exactly how much storage capacity they need.
Hansen also warned users who save corrupt files can throw a wrench into their data recovery plans.
“Your snapshot is corrupted and then potentially if you roll that snapshot up into a backup then your backup is corrupted, so there has to be something in place so that you can catch any corrupted data before it gets rolled up.”
Info-Tech does not track revenues of CDP products, but another research firm, Framingham, Mass.-based IDC, tracks the storage software market. In its Worldwide Quarterly Storage Software Tracker, released last month, IDC reported revenues in the fourth quarter of 2007 were US$2.9 billion, up 11.2 percent from the same period in 2006.
One indication of the technology’s popularity is the spate of acquisitions, Hansen said, noting storage vendors like EMC are integrating CDP into their base products. Another vendor, Double-Take Software of Southborough, Mass., bought TimeSpring Software Corp. of Montreal in December. TimeSpring’s CDP product, TimeData, now lets Microsoft Exchange administrators recover any e-mail or mailbox at any time. It also lets them classify data into groups, such as HR, financial and manufacturing.
Another recent launch is Archiware PresStore 3, a cross-platform backup, archiving and synchronization tool.
“A lot of vendors are integrating CDP into their backup solution now,” Hansen said. “It’s great if you can back something up but if you can’t get it back, what’s the point?”