Disastrous Disaster Recovery
September 22nd, 2008 | by kfinney |It seems that in most organizations, disaster recovery is an afterthought. After all, it’s not important until, well, it’s important! You don’t think about a solid plan until you actually need to exercise such plan and all too often by the time that comes it’s too late. So why does it take so much effort to come up with a plan to protect your critical systems? What are the hurdles to implementing that plan and is it actually going to work?
There is a saying that “Your backup is only as good as your first successful recovery.” I think this saying holds true time and time again because if you have not exercised your backup and recovery plan then how do you know it is going to work? Most businesses today depend heavily on tape backup. This is probably a tried and true method but, it has it’s limitations and it is often not reliable. In no way should you depend solely on tapes unless you don’t mind losing a little data or being down for an extended period of time. Tapes are susceptible to data corruption and it is often time consuming to rotate tapes, bring them from an off-site location (if you use one), catalog, load, and restore information from them. As well as the cost of tapes.
I think the biggest shortfall of successful backup plans has to do with the IT Budget. If you are fortunate enough, you have a virtually unlimited budget and you can implement whatever plan your imagination can come up with! This is often not the case and most companies have to weigh the cost of expensive backup solutions against funding for other initiatives. Security is typically at the forefront because what good is data backup if users can’t utilize your information due to a security breach? The gamut of disaster recovery tools is massive and can be something as simple as a script copying files, to full blown near real time replication suites. Arguably, the best backup strategy is one that encompasses a range of strategies and does not rely on one single strategy.
If you like tape backups, use tape backups but, include some other technology to offset the potential shortfalls of tape. A backup to disk strategy may work better utilizing the inherent redundancy of today’s SAN technology. Carve out a 1TB lun to use as a backup to disk area (or whatever size you need) and the rest is covered. Most SAN’s utilize some form of RAID, generally RAID 5, and then a feature such as snapshots to help recover to a point in time. Even further, that information can be replicated to another SAN over the WAN or just to another data center. This is a simple solution that utilizes existing technologies but, could raise the price tag depending on the number of disks needed to make this solution work. Full replication suites can be a costly alternative but, often provide the best coverage in terms of a solution that is going to provide near real time recovery and automatic failover. One such solution is Double Take which seems to be an industry leader. The range of their products can provide local recovery as well as disaster recovery and remote replication over the WAN. A number of companies offer such technology but, you want to make sure that once a failover occurs, that you are also able to failback! This is often the shortfall of some solutions.
In the end none of these solutions are worth a penny unless you can execute them efficiently and without error. This is why testing your recovery plan is so important. Relying on the fact that it “should” work is not enough to justify not testing. You don’t want to be in the situation where you actually have to rely on it and somethings not right. I know this seems like common sense but, you would be surprised at the number of organizations that actually don’t test this process. Remember, “Your backup is only as good as your first successful recovery.”
