Despite the continued lowering of costs across the board for IT solutions and IT infrastructure components, small businesses can still run into significant challenges in the design and implementation of backup and disaster recovery (DR) strategies. Current IT solutions are using more and more storage, and as business volume and connectivity grows, the need to retain transaction data, log data, and archive databases increases.
Old solutions are being outpaced by storage needs, and the 24/7 nature of Internet business is beginning to make off-site storage of tape media not only impractical in terms of restore response time, but antiquated and not meeting with the desired service level of businesses.
Fortunately, new technologies in storage such as Internet and remote backup solutions are available. Also, cloud computing, WAN fail-over, multisite clustering, and other HA and load balancing solutions are available to provide business continuity and continued availability.
Fireproof and waterproof storage is becoming affordable and practical to use as an on-site alternative to Internet and WAN-based remote backup options. But what about when a catastrophe does happen? What about after that catastrophe, with the unfortunate discovery that perhaps your DR strategy and plan simply doesn’t cut it? Hopefully you won’t ever have that situation.
Let’s take a look at what challenges exist for SMB’s in the design, provisioning, and implementation of sound backup and DR solutions.
REDUNDANCY, LICENSING, AND RELATED COSTS
It may appear at first that having redundant systems at your DR site doubles your software costs. Fortunately this is most often not the case. For many software vendors, DR systems that are inactive and turned off do not require an additional license. If only one copy of the software is going to be run at a time, the license that applies to your production system is valid for the DR system as well. This can be a boon in savings for the DR site; just be sure that you are in compliance with licensing for your products--don’t run the systems at the same time, and be sure that for the vendor in question, that an inactive (turned off) system doesn’t require an additional license.
MEDIA AND BACKUP FAILURES
Failures of backup software or technology to be able to correctly restore to a known good or prior state happen far more often than you would think. One way to avoid the catastrophe of a failed restore is to test (fully) your DR plan and validate that all the data you think you’re backing up really is being backed up correctly and can be restored properly to the desired state.
I’ve seen many cases where clients were sure their backups were taking place, yet when we would go to perform a restore, we just couldn’t get things to work. Issues like locked files, data integrity issues, backup of active databases or other information stores being problematic, and requirements for low-level replication or shadow copying that isn’t in place are all causes of these types of failures. Sometimes backup media itself can fail or the data written to it can become corrupted. Testing backups may be boring or tedious, but if you don’t do it, you’ll get burned.
BACKUP DATA ACCESSIBILITY AND CATALOGING
Full backups and system images are key parts of the backup and recovery puzzle, but not every backup can likely be a full backup. Snapshots and images take enormous amounts of space. To retain different versions of numerous changed files across many systems means daily incremental or differential backups.
So, when a restore is needed, not just one backup file from one storage location may be needed, but possibly dozens. Having a solution that makes restoring from a series of backups simple and effective is important. Finding the backup sessions you need quickly and being able to restore them easily can make all the difference when you are in a crisis.
DR SYSTEM AVAILABILITY, SETUP, AND CONFIG TIME
Having the data archived and accessible is only one part of the puzzle. The systems that will use that restored data and software configuration have to be where they need to be, plugged in, powered on, and the restores have to happen, in the right order. A DR strategy that constantly keeps the DR site up-to-date with the same versions, patches, configuration changes, and other updates (user accounts) all completely in sync with the production systems is a rare case. DR systems that are online, and being replicated to for every change in production is the ideal situation, but the most costly.
Consider the licensing and system costs mentioned earlier. Since this scenario isn’t likely, you will need to have a “run book” that describes what needs to be done, in what sequence, what administrative accounts and passwords are to be used in bringing the DR systems online, and restoring your production data.
Don’t let the challenges inherent in creating a strong backup plan and disaster recovery plan prevent you from having one. If money issues are the limiting factor, come up with the best you can with the budget you have. Then, take a look at the real costs of the downtime, and see what the peace of mind would be worth to have the ability to keep operating, or how much revenue getting back to work sooner would provide.
The opportunity cost of lost business due to a widespread problem that affected your and other businesses should not be overlooked. If you can service customers when competitors can’t, that’s a strong position for your company. Even a simple plan that’s solid, tested, and that works is better than nothing.
What do you have?
LEE CLEMMER has owned two successful IT consulting firms in the Atlanta area, and has been accepted into the executive MBA (eMBA) program at Kennesaw State University. Reach him at [email protected].