There is little doubt in an IT Manager’s mind that when it comes to saving costs on your servers or desktops, virtualisation is quite possibly the number one cost saver.
But how many people fully leverage the virtualisation tool to deliver business value? It is widely known that the time-to-delivery of a virtualised server is dramatically shorter than procuring a physical server, but how much of an issue is availability?
The phrase “high availability” often sparks ideas of clusters, application distribution services, and multiple instance scenarios, amongst other things.
In virtual environments however, this can be delivered (in VMware terms) through distributed resource scheduling and high availability tools that exist in the ESX deployment.
When used together, these tools can restart a failed system without user intervention on another piece of hardware and, most importantly, without impacting the systems running around it.
So why is this seen as a hidden benefit and not a driver for business continuity planning (BCP) or disaster recovery (DR)?
It appears many organisations do not yet fully implement these automation tools, or leave them in a semi-automatic state, thereby still relying on user intervention.
Service availability to an organisation is considered to be only required by top tier applications that generate income for the company.
These will then have hardware cluster nodes, DR or BCP hardware nodes, and replicated data, amongst other things. Little or no consideration is given to the lower end user applications that run the back end business or enable the workforce to continue with their day without interruption.
Yet, if the user community in a company isn’t working efficiently and effectively, surely they are burning money on the spot?
Availability through virtualisation is something that simply needs to be enabled to drastically reduce downtime for the lower tier applications. More than this, it should be given the visibility as a service enablement layer in the business case.
So why don’t some organisations use availability as a driver for virtualisation? The answer is they don’t believe it’s tangible or quantifiable.
We can, correct this however. Methods include improving existing service catalogues, defining recovery time objective and recovery point objective correctly, ensuring service level definition and being accountable for cost, risk and service to the application. T
his will create the self-funding required through to the business awareness of the capabilities of the tier, and the service catalogue supporting it.