Quite often, the person in charge of data protection in a corporate environment comes from a storage background. Storage folks know disk, SSD, tape, and they’re familiar with network speeds, connectivity issues, and allocation. So, when the assignment for disaster recovery, or backup, or archives, or anything to do with disaster recovery comes up, the head honcho in IT looks for someone with storage acumen.
Why? Well, it is kind of obvious, but if you’re the responsible party for putting a copy of all important data at the fingertips of the person who’ll need it back then you have to know about storage. You need to know about the right storage for immediate recovery, for secure recovery, and for long-term archiving.
Is that you? Are you the one with the storage smarts?
Drop a nugget to your fellow IT team about how virtual storage would make everyone’s life easier for data recovery, and you know how to make it happen. You’ll get more questions and rapt audiences around the coffee machines than that sports nerd who knows all the stats.
Most backup folks cower when the subject of their backup media comes up. Every month that goes by most certainly assures that recovery of data enters a higher failure zone. Storage gets old at a pace that’s alarming, and when the time comes to replace it the typical recovery decision requires that the data stays where it is. No practical method of refreshing the storage media exists without a virtual storage capability.
In the mid 1990’s, we at STORServer discovered that IBM’s backup software, Tivoli Storage Manager (TSM), was the only product to have virtual storage built into its architecture. Most of us came from a storage background, and were impressed 15 years ago with IBM’s design to organize all storage backup volumes into virtual pools. What a great concept. We assumed everybody else would eventually get there, but nobody is even close.
What exactly is virtual storage?
Let’s begin with the short-term and long-term problems. The biggest short-term problem is keeping the most immediate recovery information available for instant recovery. Backups usually head off to some oversized, cheap, and prone-to-fail media. And then it stays there. A virtual-based storage architecture allows the backup guru (that’s what we call our STORServer backup administrators) to cordon off temporary space for initial backups, or split backups off to several kinds of media types. Following that, the guru can move data easily, and even automatically, to any kind of media they want. These spaces are called pools. Pools are defined by type and with follow-on pool friendships. The pools are moveable from any media to any other media, all while allowing backup and recovery operations to take place without interference. Wow.
The biggest long-term problem in data recovery comes from the media type where the files, databases, and archive information resides. Companies store their backups on whatever current media they own. In three years the mean time between failures (MTBF) jumps from seldom to often. In six years the media may not have connectivity compatibility. In 12 years the media probably has a crust on it that eliminates the possibility for getting a tape head to read it properly, or its disk head to even spin.
Virtual storage organizes data into a volume, or bucket of space, that looks like a very large common file to a piece of media. The problems associated with such an elegant concept are fraught with peril. Fortunately, IBM tackled all those problems many decades ago when file system standards were put in place across all manufacturers. Those standards still exist today, and the perils have been shifted to the storage manufacturer’s MTBF. That’s exactly where virtual storage wants the problem, because virtual storage can shift from one media to another (crossing species, like disk to tape) whenever it wants to.
While most production storage continuously heads for the fastest and newest thing, recovery data usually gets relegated to what sounds cost effective at first, but after a few months, and a few years, is most certainly the worst performing and least reliable set of storage a company has on hand. The backup folks have no choice but to watch their older data deteriorate, regardless of the company retention requirements.
Virtual storage eliminates static locations for backups and archives. Backup administrators become backup gurus, because they can take advantage of everything on the market for the purpose that fits the recovery requirement, and when MTBF raises its ugly head they can move data to a newer (and probably cheaper) space.