Since back-up is such a regular chore for the IT department, you might expect the process to be pretty stress-free.
But at many companies, that’s simply not the case, says Jonathan Woods, a strategic architect at Logicalis. “Technical failures aren’t uncommon, and there are a bunch of other risks to consider: operator error, bandwidth constraints, and damage caused by poor handling of tapes, for example,” he says.
Fortunately, he says, there are a number of simple steps, which can make an example disaster recovery plan, that companies can take that should help them mitigate against a back-up disaster:
1. Align your back-up strategy with your wider business recovery strategy
The back-up process can only be successful if it is closely tied to the needs of the individual applications that your organisation uses to carry out its day-to-day business activities, says Woods. “For every application, you need to have a recovery time objective (the maximum time between an event and the time at which a system must be returned to operation) and a recovery point objective (the maximum acceptable time between the last available back-up and the time a disruption occurs). You also need to know which servers those applications are running on, and where they are physically located,” he says.
2. Size your solution carefully
A common mistake is to underestimate the size of the back-up solution you require, says Woods: “Make sure you do the maths properly: have you got enough tape, enough tape drives, enough bandwidth? Is the back-up server capable of handling the number of drives you require? Have you sized for the future, ensuring that the back-up solution will still be sufficient in a year or two years’ time?” And for applications that demand the highest levels of availability, he adds, back-up may simply not be appropriate – a full replication of operational data to an offsite location may be required instead.
3. Pick your media well
Tape doesn’t last forever – it tends to deteriorate over time, expecially when data has been written and overwritten to it numerous times, says Woods. Still, it is cost-effective, reliable and newer tape technologies offer greater capacities than ever before. An alternative to tape, adds Woods, is low-cost Serial ATA disk technology, which performs particularly well for full, sequential backups, but tends to be less suitable for incremental back-ups, where only data that has changed since the last back-up is captured. “Either way, pay attention to the manufacturer’s instructions for use, and ignore the expiry date given for a piece of storage media at your peril,” says Woods.
4. Don’t walk away!
Storage administrators may dream of a fully automated back-up, but the truth is that, in most environments, some degree of monitoring and human intervention will be required. “You’ve got to keep your eye on a back-up process. Never, ever assume that a back-up has been successful without verifying whether that’s true or not” says Woods. “It’s simply not worth the risk”.