Recovering your data in the event of a catastrophic failure

Have you ever sat down and really thought about how you would recover your data in the event of a catastrophic failure? You backup everything though, don’t you? Worst case scenario, you can restore everything from your backups, right?

It all sounds very simple and straightforward to get things back to “business as normal”. In practice however, this is rarely the case unfortunately. Have you thought about how quickly you could provision new servers to restore your data onto? Have you thought about the possibility of having to relocate your systems to an alternative datacentre? Have you really thought about the length of time it could take to restore your systems to a fully working state? Do you have a DR plan or runbook that details a complete list of steps it would take you to get business back to normal following every possible disaster scenario? Realistically, how long could all this take? Days? Weeks? Possibly longer? How would the business function without its IT systems?

When you stop and think about all of this in detail, it becomes apparent that in most cases a simple backup routine is not enough to effectively protect your systems from a disaster. Backups can complement a disaster recovery solution but should never be a complete substitute for one.

The good news is that virtualised systems can be a lot easier to protect from disasters than physical servers. Virtualised systems are made up of a collection of files so relocating a server can often be as simple as copying files to a new location.

VMware vCenter Site Recovery Manager not only helps us copy these files to a second location but allows us to fully manage everything which would need to be done if your primary systems became unavailable for any reason.

But Site Recovery manager is expensive, right? We would need twice the amount of hardware that we currently have. Well, probably not. Do you really need to protect those 40 test servers and get them back online immediately? Again probably not. Would you really need to instantly recover the development team’s 60 servers or could they realistically wait a few days to be restored from backups. In reality, how many of your systems are truly “business critical”? Which of your servers could the business really not function without? When you think about it like this, your recovery site could possibly just contain a couple of relatively inexpensive virtualised servers along with a fraction of your primary storage capacity, protecting only your key systems. You may choose to take advantage of “vSphere Replication”, which is included at no additional charge with most vSphere licenses, eliminating the need for extremely high speed links between sites and removing the need for identical expensive, high-end storage at your recovery site.

One of the most compelling features of VMware vCenter Site Recovery Manager is the ability to fully test and demonstrate the DR process with absolutely no impact on production systems. Testing results can be saved and used to demonstrate to the business, compliance with disaster recovery requirements. Knowing for a fact that your DR plan has been fully tested and really works is an extremely comforting thing for a business to know.

Malcolm Smith - Technical Consultant - Celerity Limited