Andrew Brewerton, EMEA technical director, BakBone Software on the merits of data centre efficiency

In some ways virtualisation is a throwback to the days when several applications would share a single minicomputer or mainframe. In between then and now, we removed complexity and interdependencies by giving each application its own cheap server. As the server population grew and floor space and energy consumption became major issues, having numerous servers was no longer sustainable. Enter virtualisation.

Virtualisation brings back the efficiency of shared hardware while keeping the technical and administrative advantages of having an individual “environment” for each application. However virtualisation is much more than server consolidation; it can make other vital contributions such as improving application availability and enhancing disaster recovery (DR) capabilities. This is because the ability to run multiple virtual machines - or VMs - on each physical server means users can replicate an application server without having to double their hardware population. However, these benefits only become apparent if done right; a flawed server virtualisation strategy can make backups harder to manage and more expensive. Also, running dozens of VMs on a single piece of hardware puts a lot of eggs in one basket. And there are risks in virtualisation being so easy to do because it allows departments or even individual users to set up VMs unbeknownst to the IT department.
A virtualised server has a hypervisor, a compact and highly efficient layer of supervisory software sitting between the hardware and the OS. The hypervisor pools the servers’ physical resources, allocates and presents them to the various guest OSs (a VM) and applications with all the software hooks and interfaces that the hardware would normally present. Each guest OS therefore “sees” a physical server but in reality is running in a virtual server; advantages include:

• isolation - if one application or VM crashes it should not affect the others
• hardware independence - users can run the same VM on any other VMware system without modification and without needing to reinstall the application or OS
• encapsulation - each VM is saved as a file and backing up, moving or replicating a VM can be as simple as copying or moving that file, and
• partitioning - which allows users to not only run multiple VMs on one physical machine, but to have a single machine support multiple different applications and operating systems.

A key element of any virtualisation strategy is data protection; virtualised servers can be much more productive than physical servers, yet this makes protecting this data even more important. One way to deliver data protection in a virtual environment is to move the backup and recovery load off the application server and into the storage tier; many tools are available to protect virtual servers today. More is needed though because virtualised infrastructures work differently from physical ones and hence need different data protection techniques.

Virtualisation is proving particularly powerful when it comes to business continuity (BC) and DR; BC is all about maintaining application availability and delivering an uninterrupted service and part of that is DR, where users might need to rebuild a failed, damaged or stolen server, or perhaps switch an entire operation to a secondary datacentre.

While it is true that server virtualisation can make the system administrator’s life simpler and more cost effective, it will also add complexity in a number of notable areas: as departments and end users set up their own VMs, tracking and discovery will become major challenges. Management and protection will also be challenging: managing and protecting 40 VMs is one thing, but 4,000 are a different ballgame altogether. Data protection will need to be planned differently too and possibly include newer technologies such as CDP/RDP (Continuous/Real-time Data Protection) and replication. Server virtualisation will also affect the storage infrastructure. If a user wants to migrate VMs freely they will probably want storage virtualisation too. To help here they should look at technologies such as thin provisioning and de-duplication for example.

The logic of server virtualisation is inescapable; it can make servers much more efficient but it can also do a whole lot more. It is just a case of understanding the business and IT needs and building a virtualisation strategy that meets these requirements.

• www.bakbone.com