With news of the recent earthquakes in Christchurch, New Zealand, my mind jumped not only to the poor individuals involved in the disaster, but also to the technology devastation that will have occurred. What do I mean by that? The fact is the majority of SMBs will not have planned to deal with this type of event.
Many SMB IT consultants install backup software to cater for the loss of data on their clients servers. However, most do not ensure a regular test of the recovery of this data is performed. You can backup as much as you like, but unless you periodically test the restoration process you do not know for sure that your backups are good.
The simplest form of test is where you restore a set of files from a recent backup and then verify that those files are indeed valid and can be accessed. This is something that as a reseller we can document for our clients and give them a simple procedure to follow. I’d recommend that you do this on a monthly basis and once the procedure is setup it should take the customer only 10-15 minutes. It would normally only take a reseller an hour at most to document this procedure for the client and then put in place some form of logging so that the client can prove they have tested their backup/restore strategy.
A more complex form of recovery test is where you take the most recent backup of your server and restore it to different hardware. This type of test has a higher risk of failure depending on how you backup your systems. It can also be harder to perform for the reseller as you will need hardware of similar or higher specification to perform the restoration process. I’d recommend that you perform this type of recovery every 12 months or so.
All the recovery environments above really only talk about small/simple single server failures, but what if you’ve had your entire computer room wiped out because a wall has fallen onto it? That’s the type of thing most likely being faced in New Zealand this week. The problems they will face start with being able to source hardware in a way that is both fast but also not costing an enormous amount. Insurance companies will typically only pay out claims months after the event which means the client is forced to finance this now either from their cash flow or via some form of bank loan. Then there is the challenge of actually procuring the hardware quickly.
All of the above can be avoided with proper planning.
Check your insurance policies and find out who long it will take for the insurer to pay in event of a massive claim like the ones in New Zealand. If you can’t afford to wait that long then you best look to save a portion of your profits and cash flow for an event like this.
Put in place procedures for your clients to check their backups on a routine basis. Make sure you document the fact that the backups worked and what was tested.
Put in place more complete disaster recovery procedures. Invest well in software that allows you to test recovery of your entire environment in as short a time as possible. Make sure that you then test that periodically and again document this and advise the client.
Hopefully, the issues in New Zealand will be not as severe as the media reports and we don’t have businesses going out permanently due to the lack of tested recovery procedures. Plan for disaster now and you can avoid suffering as much as others will.