Following any disaster, IT administrators have to do three things: recover data, re-host applications and reconnect any networks. When employing high availability, these processes are automated with self-healing networks, reconstitution of network processing and a process to restore data, said Jon Toigo of Toigo Partners International.
But such efforts are costly to use for all data and every application in your environment, Toigo said. And there are other potential problems. "If it were [possible], the old ideas about recovery and business continuity from a hardware implementation standpoint would be history ... unfortunately, all the hardware and software has to work for that to be true," he said.
Toigo said determining whether data is critical depends on how long an organization can go without access to it. He said that data inherits its criticality from the application that it supports. "Not all of your applications require high availability. You're wasting money throwing high availability at an application you don't need access to for 30 days if there's an interruption," he said.
For data that doesn't require high availability, Toigo said, tape remains a good option for storage. "Tape makes a really good modality for protecting that kind of data … in fact, I would say, do not use a cloud if they cannot give you tape," he said, adding that tape can serve as a good archive for data that doesn't get accessed frequently. He noted that he has a concern about cloud providers that don't use tape for recovery purposes.
"Now, if my cloud provider can't deliver … a tape recovery service where they stick the tapes in a box and ship it to me via FedEx, I'm not doing business with them. It's really just that simple," he said.