What is disaster recovery?
- Disaster recovery
- Disaster recovery is the documented plan for restoring systems and data after a serious failure, covering what gets restored, in what order, and how long it should take.
What disaster recovery means in practice
Two numbers define any plan. How much data you can afford to lose, and how long you can be down.
Choosing them honestly is the hard part, because every business wants zero on both and neither is cheap.
Backups that have never been restored are the standing joke of this field, and the joke keeps being funny because it keeps being true.
For a phone line the plan is unusually simple. Where do calls go when the usual destination cannot take them.
Ordering matters more than speed once something has actually failed. Bringing the phone back before the reporting dashboard is obvious in advance and surprisingly easy to get wrong at six in the morning.
What people get wrong
A flooded server closet at 6:10 am
Say a pipe bursts over the weekend at your six-person accounting office. At 6:10 on Monday you find the closet holding the file server under an inch of water. Your plan is one page taped inside a cabinet door.
Step one, at 6:15, has you sign in to the carrier portal from your cell and forward the main number to it. At 6:30 you text the staff group to work from home on the cloud copies. By 8:00 you've called the IT contractor whose number is printed on the page, because nobody remembers it otherwise. The last step restores Friday night's backup to a spare laptop, which the page says took three hours and ten minutes the last time someone tried. Client files are reachable again by 11:30. You lost Saturday's scanned receipts and nothing else, because the page held real phone numbers and a measured restore time.
A restore drill you can run this month
Pick one folder that matters, such as last year's client returns, and restore it from backup onto a different machine. Time it. Open a few files to check they're readable and not merely present. Most first drills turn up something awkward. Maybe the backup password lives in the head of someone who left, or the backup skips the folder you assumed it covered. Sometimes the restore takes nine hours when you'd guessed one.
Write the measured time on your plan. Then do the phone half. Forward your main number to a cell using the carrier portal, call it from another phone, and turn the forward off again. That takes five minutes and proves you still have the portal login.
Skipping the drill has a simple cost. An untested backup is a guess, and you'd be checking that guess on the one morning you can't afford a wrong answer.
How GreetKeeper handles it
GreetKeeper offers no service level agreement and publishes no uptime figure, because we have not measured one honestly.
The practical protection is the fallback you set at your carrier, which keeps working when any individual system does not.
Set it to ring your own number, and test it once so you find out it works before a morning when it matters.
Disaster recovery questions
What is the minimum plan for a small business?
A written fallback number at the carrier and a tested backup of your customer data. Those two cover most of what actually goes wrong.
How often should it be tested?
Once or twice a year is enough to catch the big surprises. Untested is where almost every failed recovery starts.
Who decides the targets?
You do, by deciding what an hour of downtime costs. Vendors quote numbers; only you can say whether they are acceptable.
Related terms
Hear it take one of your calls
Two minutes, your own scenario, no card.