Implementing virtualisation through Hyper-V
Server virtualisation can provide huge benefits to organisations of any size, as this feature explores
In this three-part series we’ll discuss how server virtualisation can provide huge benefits to organisations of any size.
In the first part we’ll show how you it’s possible to save costs by implementing a flexible and scalable virtual infrastructure. We’ll also list some of the huge advances Microsoft has made in Hyper-V for Windows Server 2012.
The benefits of virtualisation
Discussing all of the benefits of virtualisation could easily fill the whole of this series, so we’ll confine ourselves to just a few main points. With recent advances in virtualisation, even resource-intensive services such as SQL can run in a virtual environment with almost the same level of performance as running direct on a physical server. This means that most servers are likely to be candidates for virtualisation. But what does a company gain by virtualising its servers?
Cost savings are often the driving force behind changes in infrastructure, and virtualisation can score highly in this area. Take the example of a network running three ageing physical servers. There are two choices: either buy three more servers to replace these, or buy one server and run three virtual servers within it.
Often physical servers don’t utilise their resources fully, and so even a relatively low-cost server may be able to run all three virtual servers without problems. A single physical server is also likely to draw less power than three separate servers, so cost savings can be increased further.
Even a relatively low-cost server may be able to run all three virtual servers without problems
Virtualisation provides a flexible framework that can meet the ever-changing needs of a business over time. Providing more memory or disk space to a virtual server can be done with a few clicks of the mouse. Once all the available resources of the servers have been used, simply buy another one and move some of the virtual servers across to it.
This can all be done without even needing to shut the virtual servers down, so users wouldn’t even be aware of the change. If at a later stage there’s a need to add a shared storage product, it’s also possible to move the virtual servers across to that without any downtime. Compared to the complexities of moving services between physical servers, it’s not hard to see why virtualisation is a much better solution.
In an ideal world each server would run one or at most a small set of services, so web servers wouldn’t also double up as a file, SQL and remote desktop server. With a limited number of physical servers this can become impossible. A large number of services and applications on a server increases the chance of conflicts occurring, and also makes troubleshooting problems a more complex task. Also, if there’s a need to reboot a server, then all the services will be affected (and, of course, all the users using those services.
Contrast this with virtual servers with each providing a single service or small set of services. Conflicts are less likely to occur and individual servers can be rebooted if required. So if a web server is about to crash it can be rebooted on its own, leaving all the other services available.