• Press Center
• Press Releases
• Media Resources
• Media Contacts
• In The News
• Awards
• Submit Feedback

Newsletter

Press contacts:

Bill Taylor-Mountford
CRN
September 01, 2009

Backup to the fore

With more and more organisations adopting virtualisation technologies, enjoying cost savings and simplified server management in datacentres, Virtual Machine (VM) backup and recovery has become a major challenge to IT administrators.  

Virtualisation allows organisations to deploy multiple logically distinct VMs, each of which runs its own "guest" operating system, on a single physical "host". Llike any physical server, virtual servers contain a constantly changing set of user data, settings, and applications that must be protected. Simply backing up a physical server cannot reliably capture the complete state of a running VM, therefore cannot always ensure recoverability. In fact, VMs are even more vulnerable due to the nature of the virtualisation technology – in addition to hardware loss and failure, compromise of either the host operating system (OS) or the guest OS will cause data loss and impact productivity.

Backing up from the inside

To minimise the interruption to the business when disaster strikes, organisations have to re-evaluate their backup strategies and provide VMs with the same safeguards as physical servers. With so many options on the market, choosing a solution that suits your needs is crucial.

First of all, a good backup solution should support all the major virtual platforms, such as VMware, Microsoft, Citrix and Parallels, so that you have the flexibility to deploy the best platform according to the environment. 

Secondly, the solution must back up VMs from within each VM to provide the complete protection – treating each VM like a distinct physical server that needs to be backed up independently.

Disk imaging is proven to be the best path to take because it reads the sectors and creates an image of the entire VM, including all data and operating system components. It can even backup and restore individual files and folders on your VM while online — keeping users productive. Not only does it give you peace of mind knowing that your entire system is fully protected, but also enables you to restore the VM in minutes, not hours or days, in any disastrous situation.

Some organisations even use the created disk image for migration to and from physical and virtual environments, beyond just disaster recovery. Uses include staging new servers before they go live, testing the performance impact of adding a new application to a server, and load balancing to easily move the assets between servers, which adds new value to the business.

Agent-less VM support

One thing to be aware of is that most of the backup solutions require organisations to install individual agents on each VM to be able to perform backing up from the “inside”. Since there could be up to 20 or 30 VMs on one physical host, this requirement creates extra work for IT administrators, and more importantly, costs much more in software licences.

Recently, a few key industry players introduced agent-less VM support, allowing VMware or Microsoft based virtual servers to be backed up without installing individual agents on every virtual machine. Users only need one single agent to protect all hosted VMs, and VMs are managed the same way as physical machines. This innovative approach can dramatically reduce the resources and costs required for VMs, as well as simplify the way that VMs are handled and maintained. 

Another important factor to consider is licensing. Many solutions charge a licensing fee for each VM. As a result, it can become extremely expensive for an organisation to build, test, migrate and manage a VM infrastructure. A more flexible and reasonable licensing model is to charge the software per host. Users can install unlimited numbers of VMs on one host, and be charged at a flat rate. This way, an organisation can accurately predict software costs with no worry of blowing the budget as the data centre expands.

Duplicate data prevention

Backing up the content on VMs, like any other backup, can result in the existence of duplicate data being on the organisation’s network. This existence of duplicate data leads to a greater management burden and higher operational costs, not to mention the excessive storage space required. Exponential data growth becomes a major concern in an organisation’s day-to-day operation.

The most effective measure to tame this relentless data growth is data deduplication. This technology can cut storage requirements by up to 90 per cent by eliminating backup copies of redundant data. It’s a great tool to reduce storage costs, but the technology itself is not cheap.

Traditionally, data deduplication was provided by storage hardware vendors as a function of their hardware solutions. The technology being offered can be extremely expensive and beyond the reach of most organisations.

It’s good that we are seeing more and more software deduplication solutions introduced to the market. These solutions are usually vendor independent, giving organisations the flexibility to choose the storage devices suitable for their environment. But not all software solutions are affordable. Purchasing the licences by storage capacity or CPU can still end up with a huge investment, whereas solutions priced per machine are more reasonable and can genuinely reduce the operational costs. It will be even better if the deduplication technology is offered by backup vendors. By integrating data deduplication into the backup process, you can manage your data through one single solution.

In addition to the costs of the technology, you also need to be aware of the options you can get out of the solution, for example, is it file or block level deduplication? Is the deduplication performed at the source or the target? All the above would be good.