If your business can't run for days while you rebuild lost data, you need these ten things in place before anything goes wrong. A disaster shows up at the worst possible moment, and the time to figure out your response is not while your server is dead. Get these set up now and you can be back up and running fast.
1. Write the plan down
Just thinking through what happens when your server melts down or a fire takes out the office gets you back faster. A plan doesn't have to be fancy. At a minimum, list the disasters that could hit you and, for each, a step-by-step process: what to do, who does it, and how. Include contact information for every provider you'd need to call.
Decide how long you can actually be down. If the answer is a few hours, the plan has to get you back inside that window. One way to hit it: virtualize your server so the office keeps working off the virtual copy while the real machine gets repaired.
2. Get a pro on the recovery
Recovering data after a disaster without help is a great way to lose it for good. One wrong move during recovery can erase your data permanently or stretch downtime into weeks. Work with someone who has done both sides of this: building contingency plans, so you have a framework to restore from, and actual data recovery.
3. Plan how you'll communicate
During an emergency, information keeps people calm and keeps work moving. Getting accurate, timely updates to employees, customers, and vendors protects trust, safety, and productivity. A communication plan written ahead of time means you're not inventing it under pressure. Yours should cover at least the following:
- A current emergency contact list.
- A clear chain of command.
- More than one way to reach people, by email, text, or phone, so you can keep employees, customers, and vendors updated.
- Written emergency procedures.
- Copies of key contact records and documents stored off-site.
4. Automate your backups
If a backup depends on a person remembering to do something, it will fail. Human error is the number one cause of data loss: someone forgets to swap a tape, or the backup was never set up to run right in the first place. Automate it so it runs like clockwork, with no one in the loop.
5. Keep a copy off-site
Always keep a recent copy of your data off-site, on a different server or storage device. Onsite backups are useful, but they go down with the ship. If your server gets stolen, flooded, burned, or hacked, a backup sitting next to it goes too.
6. Set up remote access
Remote access keeps you and your staff working when you can't get into the office, and it's a convenience the rest of the year too. It also lets your IT staff or consultant reach the network during an emergency or for routine maintenance, without driving in.
7. Image your server
An off-site copy of your data is good, but data has to be restored somewhere to be useful. Without the software disks and licenses, standing your applications back up (Microsoft Office, Creative Cloud, your database, accounting software, and the rest) can take days even when the data itself is sitting right there.
A server image is an exact replica of the machine. You copy it straight onto another server, which saves a huge amount of time and money getting back online, and you don't lose your preferences, configurations, or favorites. Ask your IT professional whether this kind of backup fits your setup.
8. Document the network
Network documentation is a blueprint of the software, data, systems, and hardware on your network. Have your IT manager or consultant put it together. It makes restoring your network faster, easier, and cheaper, and it speeds up everyday repairs because technicians aren't hunting for where things live or how they're set up. If disaster does strike, it's also your record of exactly what you lost for an insurance claim. Keep a printed copy with your disaster recovery plan.
9. Maintain the system
Most downtime never comes from a fire or flood. You're far more likely to lose data to a virus, a worm, a hacker, or plain human error. So keep the network patched, secure, and up to date. Watch hardware for wear and software for corruption too, an overlooked failure that can wipe you out just as fast. Replace or repair aging gear and software before it quits on you.
10. Update, test, repeat
A backup you've never restored is a guess. A 2017 survey by Forrester Research and the Disaster Recovery Journal found 43% of companies run a full test once a year and 19% run one at least twice a year, while only 14% keep their plans continuously updated and tested. Have an IT professional run a test once a month to confirm your backups work and your system is secure. The worst time to test your parachute is after you've jumped out of the plane.