Disaster Recovery, in short, is the process of ensuring recovery of technology infrastructure vital to an organisation after a disaster. This, to Plan B, means the ability for your employees to resume productivity after a disaster has rendered your IT system inaccessible.
There are several disaster recovery solutions available, and which one you choose will depend on how long you can afford to be without your IT system, and how much data you are prepared to lose in the process.
The most basic of disaster recovery solutions is the backup – whether this is a tape or online backup, its purpose is purely to copy your data and store it in case you need to rebuild your systems. Although it has the benefit of being inexpensive, if your business suffers when your IT system is down then this isn’t the best option for you. Think of how long it takes to build a network and this is what you’ll have to do in a recovery, all without planning. How long does it take to buy a new server and then load it with software? Recovery in this way can take days, and with 19% of recoveries being unsuccessful* then you may also lose a lot of data as you trawl back to find a backup copy that works.
A step further forward there are companies that will provide you with a backup and recovery solution. It means that in the event of an IT failure they will recover your systems for you instead of you having to do it. This has the advantage that recoveries can start immediately as the hardware is to hand. During a recovery, however, there will be lots of errors which will need working on with your IT team so it will be labour intensive and complex. Often it can take many hours to get a new server with your backed up data working again, and if you have a number of servers then you need to multiply the time for each server. Some companies will give recovery objective times of 3-4 hours but in reality it can still be a lot longer if you have a larger network. It’s best to talk to companies about recovery times based on your system.
Then you come to solutions which offer you high availability. This means your IT system can be accessed quickly in the event of an IT disaster and there are two main methods of doing this.
a) Pre-recovered virtual disaster recovery – now possible thanks to the cloud. Your IT systems are recovered each night onto a virtual platform and tested. It means that your recovery success is guaranteed and the recovery time is really fast as it’s just a case of booting up your system. Automation enables this at a low monthly cost, meaning it’s affordable for companies that couldn’t afford server replication. Fifty servers can be recovered as quickly as one, so its price point and performance means it fits well with mid-market companies that take their IT availability seriously.
b) Server replication. For companies that can’t afford to have any downtime, not even minutes, then synchronous replication is the only way to go. Suitable for banks dealing with high levels of transactional data, it involves buying and maintaining a spare set of servers. This is very costly up front, and also from a maintenance point of view with regular testing being required. It does,however, allow you to experience an IT disaster with no downtime and no data loss, as you will automatically failover to the synchronised replica. In contrast, asynchronous server replication is no longer considered a good buy as the Pre-recovered virtual options available offer faster recoveries at a cheaper price.
For more information on the cost vs performance of disaster recovery methods download our whitepaper.
By Beth Baxter