By Deni Connor Virtualization has moved into the mainstream IT infrastructure. Hardly any organization exists that is not virtualizing their x86 servers to save floor space, reduce energy consumption, cut hardware costs and create a disaster-recovery scenario for them. One organization I talked to has a new IT mantra – “If it can be virtualized, then do it.” It seems that more and more, that mantra is good enough motivation to virtualize servers. In spite of the popularity of server virtualization, there are downsides. Server virtualization has resulted in a sprawl in the numbers of servers virtualized – servers are so easy to virtualize that the number of virtual machines that exist can easily expand and contract as necessary – this creates a problem for management and protecting these machines from disaster. IT managers may not know how many virtual machines they have to protect or what the configuration of each is. That’s not a bad problem to have, just one you need to watch out for – which makes the idea of management tools to manage virtualization sprawl more important. The most common motivator for server virtualization is disaster recovery and how we protect the virtual machines we create. There are a number of ways ranging from placing a backup agent on each VM itself, protecting the virtual machine as a whole by protecting the entire physical machine and all its VMs. There are also as many ways that qualify as disaster-recovery solutions – capable of getting data back fast so a business can get on its legs again. Users are virtualizing application servers and Internet servers – servers running DNS, Web, Active Directory and DHCP servers. They are virtualizing Exchange and SQL environments and front-office applications such as Word, PowerPoint and PDFs. They are using dual-core and quad-core processor servers to virtualize their environments. Those that use shared storage are adding host bus adapters and N-PIV to create more reliable, available and flexible environments. They are storing images of VMs on their storage-area networks so they can recover them easier and more quickly. For whatever reason you have virtualized your servers, you still need to manage them just like the physical servers in your environment. You need tools that help you estimate how many virtual machines a physical server can support; tools that help you discover the VMs out there and what their individual configurations are, and disaster recovery and data protection tools that let you ensure that whatever you virtualize is protected correctly. A whole arsenal of tools are needed in this brave new virtualized world.
Virtually here
Opinion
Nov 24, 20083 mins




