Let’s cut to the point. No matter how good you think your disaster recovery plan is, it’s useless unless you test it constantly. It’s no good testing it when you first implement it. It’s no good testing it annually. The number of businesses that have an outage, go to restore their systems and then find they’re unable to is unacceptable. That needs to be a non-existent problem.
Take Delta, who topped IT headlines in 2016 who had an outage that resulted in a global computer failure. More than 2000 flights were cancelled and tens of thousands of passengers were left stranded. All of this because of a problem that could have been prevented with proper disaster recovery testing.
The firm discovered that around 300 of their servers weren’t connected to a source of backup power. The lack of connectivity to the backup power contributed to the incident for sure, but there’s a deeper problem there. Had Delta actually thoroughly tested their disaster recovery progress regularly, the backup connectivity issue would have been flagged far earlier.
Storage expert Brien Posey analysed the situation and made a number of recommendations. Firstly, disaster recovery must be an ongoing progress, not a yearly event. How often you want to test it can vary on your data setup and the size of your business – consider monthly at least.
Secondly, try to find any single points of failure that will bring the entire process down and fix these with immediate priority. Failure cares not for convenient timing; if you discover a flaw, rectify it without delay.
Thirdly, have a secondary data centre that can automatically handle any critical workloads that fail on the first. Failure can always happen and you need to have a contingency plan in place should that happen. Don’t let the recovery process be held up by a non-functioning data centre.
Before conducting the actual test of your disaster recovery systems, you might want to conduct a dry run. Bring in all the key departments to ensure that everyone understands what is expected of them and the facilities are prepared. Be willing to stop during the process if something goes wrong, make notes, learn from these errors and then run the test again. Don’t give up until your disaster recovery plan is flawless.
An emerging, affordable market is disaster recovery as a service, with many companies launching products of their own. These can be useful for businesses who don’t have the resources to control their own disaster recovery plans, but understand its importance. Decide if you need it, then work out what you need to recover, the timeframes you’ll need to stick to, along with considering your security requirements.
If you don’t think disaster recovery is important, consider the fact that more than 4000 ransomware attacks occurred daily in 2016, which was a 300% increase over the figures from the previous year. Protect against ransomware by having a strong disaster recovery process to allow you get back to a workable point in time, minimising the outage and reducing a hack’s impact on your business and its data.
Your Recovery Plan is Useless Without Constant Testing
No comments yet. Sign in to add the first!