– Adam Stuflick, technical sales lead at Novastor (www.novastor.com), says:
First you need to determine if in fact your disaster recovery plan is up to date. Have you added new hardware or new software that your current disaster recovery plan does not account for? Next you need to take the requirements for your DR plan and talk with the various groups that it affects and confirm that your restore time and restore point objectives are in line with their expectations. You then have to take these requirements, change them into realistic requirements with the budget you have. Then you need to evaluate your different options that fulfill your new realistic requirements. Test your different options. Decide on a solution and implement it. When you are doing the implementation, it is particularly essential to document the implementation so you can reference it when a disaster happens, when you add new hardware or software, or when you need to train someone else to deal with these disasters.
In geographically diverse data centers the best generic disaster recovery plan is to replicate the data between the data centers, have one of the data centers be the root recovery center that you take snapshots of the replicated data in, and then as a way for a quick restore time objective either implement an internal cloud based backup or external cloud based backup. This way if you need quick access to certain files you can quickly go to the cloud based backup. If an array or data center is totally wiped out you will have the replicated data in another one of your data centers with next to no time loss in data. If something even more disastrous happens and a large chunk of data is corrupted and that is replicated, you will have the snapshots at your root recovery center that you can quickly restore large amounts of data from.
How can SMEs handle…
…recovery time objectives?
You need to take the requirements for your DR plan and talk with the various groups that it affects and confirm that your restore time and restore point objectives are in line with their expectations. You then have to take these requirements, change them into realistic requirements with the budget you have. Then you need to evaluate your different options that fulfill your new realistic requirements. Test your different options. Make sure that the affected groups are ok with your recovery time objective and restore point objectives. Decide on a solution and implement it.
…changing requirements?
Every time a new piece of hardware or software is introduced into the environment there must be a check of the disaster recovery plan. How does this new item affect the DR plan? Does this affect the restore time or point objectives that have been laid out? What new challenges does this new item in the DR plan present? Did I update the documentation about the implementation of this new item in the DR plan?
With the new options cloud based storage is giving end users of all sorts of different options for storage, be it production storage or archival storage. With these new options it gives IT administrator much more flexibility to really develop a disaster recovery plan that really covers all facets of their infrastructure. With a combination of tape backup, hardware replication between sites, and a cloud based backup solution you can have a very flexible and complete disaster recovery solution.
