Venyu Blog

Now that scientists in the UK have appeared to solve the perennial debate on which came first the chicken or the egg, perhaps one of their counterparts in IT might take a turn on the most recent dust-up over whether Server Virtualization created online-backup-as-a-service (e.g. cloud computing) or whether online-backup-as-a-service’s natural result is, in fact, Server Virtualization.

In this corner, and supporting the virtues of, well, virtualization, the online pub ServerWatch that recently posted an entry “Top 10 Server Virtualization Myths – Debunked” that discusses, according to its writer Ken Hess, the “10 myths [that] perpetuate suspicion and paranoia for those considering a switch to a virtual infrastructure.”

Among the myths debunked (the complete list is available here):

#1. Server Virtualization Is Too Expensive

Often, the primary goal of virtualization is to save money by leveraging high-end hardware using multiple virtual systems to spread the costs among those systems. Virtualization will save money, if executed correctly. Correct execution involves careful planning and prudent technology selection.

#4. Server Virtualization is Difficult to Learn and Support

Virtualization leverages standard skill sets from enterprise support staff: Linux, Unix and Windows services change little in a virtual infrastructure. The skills required to support a physical environment easily transfer to the new virtual environment. Virtualized OSes behave much like their physical counterparts.

#8. Lengthy Recovery Times for Virtual Machines

Like any well-executed backup and recovery plan, virtual machine recovery depends on the people who implemented the plan. Virtual machine restore procedures offer quick recovery compared to that of the fastest tape-based system. Alternatively, virtual machine recovery requires a simple replacement of the damaged virtual disk files with undamaged ones. Service restoration can take as little as a few minutes. Some administrators choose to use traditional backup technology with virtual machines that may prolong this effort. However, that choice has nothing to do with virtualization itself.

#9. Server Virtualization Leads to Server Sprawl

Server sprawl refers to the practice of creating virtual machines that consume virtual host resources with low return for those resources. It is neither a natural progression nor a result of using virtual infrastructure. Server sprawl has more to do with administrative control and policy breakdown than the technology it’s based on, and it happens just as often in physical environments.

Mind you while none of this actually “solves” the cloud computing vs. server virtualization debate, it’s an interesting and elaborate argument to support the benefits of virtualization as a way, absent cloud computing, to contain server sprawl and re-allocate resources in a way that’s optimized for the needs of the business.

And to bring this blog posting full circle, it turns out the chicken came first.