Disaster does strike

Although you plan to avoid disasters they will always happen, usually from a combination of circumstances that you have not anticipated or thought extremely unlikely.  What is more there may be some additional twist that will make sure recovery is not straight forward.

A week ago one of our customers lost several key servers that needed to be restored so that the business could get back to normal operation.  I am pleased to report that ShadowProtect from StorageCraft did the job and we were able to bring the servers back on different hardware, but it was not an easy ride and the experience reinforced several points.

1 Test your disaster recovery process so that you know

• How long it will take
• The equipment you will need
• There will be no critical decisions to make
• You can be confident because you have done it at least once

2 Restoring large quantities of data can take a long time.  Business data is continually growing; do you really need to restore 1.5TB of data for the system to work? Will it be 3TB next year?

3 If you lose your servers you will need alternatives to restore to. In this case there was almost sufficient capacity available; only one new server had to be provided.

4 There is nothing worse than going through an 18 hour restore process for it to fail because of some equipment incompatibility and you have to start again.  On this occasion we hit a 2TB to 3TB disk compatibility issue; we had to do some fast thinking and start again.

5 The benefit of practice and following a documented process cannot be underestimated when the pressure is on and people want to know when they can get back to work.

Everyone will agree that practising the DR process is a good idea, but they do not always want to pay or make time for it.  If you are responsible for this do not be put off, you need to do everything you can to reduce stress and ensure success in the event of disaster.

There are many things that I will look at in more detail over the coming weeks.