If you do your disaster recovery plan right, it should not fail. If it does, then you have overlooked something in the planning process. There are ten major reasons why disaster recovery plans fail – adding a second disaster to your first. As human error or hardware failure are far more significant problems than natural disasters, your disaster recovery plan needs to be as close to foolproof as possible. Here are the ten things to consider:
Too many companies write up a detailed disaster recovery plan – and then don’t communicate it properly. In many cases, this is because management are putting in a plan because audit or their insurance company require it, without any actual buy-in. The plan should be available to everyone on staff, with multiple copies at multiple locations. And remember that the time you need the plan is likely the time you don’t have a computer (or power) and store multiple hard copies in the office and in the homes of key personnel.
Plans need to be tested, updated, and discussed regularly. A plan sitting on a shelf may not include vital details, may have incorrect phone numbers, etc. The plan should be tested as often as practical – usually, this means once or twice a year. Documentation should be updated any time a relevant change happens.
Too often, backup servers are considerably lower powered than active servers when, in fact, they need parity. Make sure you can continue production at a good level while recovery operations are completed.
Remember that people need to be able to follow these instructions at 3am, with no power, while the wind roars outside. Employees doing different tasks may not be able to directly communicate with each other.
Consult with employees about what really needs to be in the plan. Make sure you have a list of everything you need to resume functions and that the plan covers all of it.
Too many companies only keep enough fuel to run their generator for a day or two. In some places this is alright, but if you are working in an earthquake or hurricane zone you need to have enough fuel for several weeks. Your generator also needs to be powerful enough to keep essential functions going.
Maybe there is not actually enough space on your backup server for all of your data. The middle of a disaster is also not the time you want to discover that your backups are corrupted or your backup system hasn’t actually been pulling anything for six months. Test and check backups frequently to make sure they are current, clean, and cover everything you need. Also, make sure at least one set of backups is far enough away not to be hit by the same disaster.
Again, consult with your senior employees – they may know better than you what you actually need to do. Make sure everyone is on the same page about what the most critical functions are. This is another reason plans need to be tested properly.
A plan is one thing, but do you know who’s going to fill in for missing employees? Do you have people cross-trained so that essential employees can keep things going when everyone else evacuates?
As in, your tests bear too little resemblance to the reality of the aftermath of a disaster. In addition to testing often, try to do drills that focus on the specific situations you are likely to encounter and vary your drills to simulate different problems – such as a hurricane or a terrorist attack.
If you keep these ten things in mind, then you have less chance of being left in the lurch when your disaster recovery plan turns out not to actually work. If you need more help, contact Otava for assistance with cloud backups and disaster planning.