(The Hosting News) – Technological evolutions are driven by the fact that no matter how sophisticated a technology’s defence, security, productivity or efficiency measures are today, there’s always a risk that they’ll be unsuitable tomorrow.
It is this simple notion that has driven virtualisation to become one of the top disaster recovery solutions in business computing.
A catalogue of operational challenges with client-server based applications and traditional x86 server deployments have made enterprises question if relying on their own physical hardware alone is really the way forward in supporting their business continuity measures.
For example, is running one application per server really efficient? Are the power and cooling requirements of these servers justifiable; considering they run at full power no matter whether they are in use or not? Are the costs associated with running – or rebuilding – such complex, dedicated systems really worth it?
Most importantly, can an organisation afford the possibility of losing this business-critical resource for long periods of time, should the worst happen?
Where does Virtualisation fit in?
Virtualisation has become the answer to many of these questions, as it is generally agreed that it has the ability to save businesses money, reduce energy costs and improve the timing/quality of recoverability.
It is this point of recoverability – the ability to restore a deployment to the point at which a failure occurred – that is making the most waves within organisational business continuity. That’s because where it might take days or weeks to get physical (hardware-based) servers back up and running, virtualised environments take mere hours.
A 2010 Aberdeen Group Survey showed that organisations that had moved at least part of their data storage to cloud-based contingencies, such as virtualisation, recovered from downtime events almost four times faster than those with no formal cloud storage program.
This meant that the average downtime for mid-sized businesses using cloud contingencies is around four hours, rather than 16; while for large businesses it is around 7.7 hours, as opposed to 30.8.
Using virtualised environments, whereby a company’s physical data is mirrored elsewhere, a company can also reduce its energy use. That’s because there is less reliance on using an energy-sapping physically-stored data server as a primary resource for its backup needs.
Organisations such as virtualisation specialist VMware claim that virtualisation could in fact create savings of 50 – 70 per cent of a firm’s overall IT costs as businesses begin to consolidate their resources, reducing overheads in the process.
So what do businesses need to do?
The requirements of a business venturing into virtualisation are few and far between, making it a disaster recovery option that’s near-enough available to anyone.
For example, virtualised systems are independent of their hardware and therefore can be implemented in a way that gives power to the user. Virtual components (e.g. CPU, network card, SCSI controllers) can be scaled accordingly to the desired task, even if the actual physical makeup of the host machine carries much greater processing power. To the same end, this means that virtual machines on the same server can run different kinds of operating systems (Windows, Linux etc.).
The relationship between disaster recovery and virtualisation is even more unique because virtualised environments (or ‘machines’) can work in isolation, independent from one another. This offers the ability for three virtual machines to pick up the load in place of another that may have crashed, for example.
Knowing just what to virtualise is a different matter. In a perfect world, the hardware configuration for a disaster recovery solution should mirror the configuration of the primary system in the data centre. However, storing a second, identical copy on another server is an expensive proposition as it will need constant power and potentially as much maintenance or supervision as the primary server.
With virtualisation on the other hand, it’s possible to have an image – or virtual copy – of the primary system. This will guarantee users 100 per cent recovery in a much shorter space of time, with less manpower than it takes to rebuild a traditional data server in the event of a disaster.
Weighing up the options
Like any disaster recovery solution, the potential savings facilitated by virtualising a system must be weighed up with its total cost versus its initial outlay.
While in some cases it may well work at a ratio somewhere akin of 1:1.5, the downtimes associated with manually restoring a physical hardware system are arguably more time consuming and expensive. It’s for these reasons – coupled with the fact virtual backups can be kick-started remotely – that virtualisation is taking over as the new solution to tomorrow’s problem.
For information from Peer 1 Hosting, please visit: http://www.peer1hosting.co.uk/