Is Capacity Planning Like Using a Magic 8-Ball?

Capacity planning has become an absolute necessity in effectively managing your dynamic virtual environments. Virtualization itself presents fundamental challenges in understanding the impact of resource sharing such as CPU, memory, disk and network across both virtual and physical environments.

Best-in-class capacity planning solutions, such as in vFoglight, provides the necessary tools and technology to properly identify capacity trends to effectively plan for growth and avoid costly unplanned outages. But beware of other virtualization capacity tool vendors have been making claims for predictive capacity planning, for example, “You can fit 22 more VMs here”. Vizioncore questions the foundational and technical basis for this type of analysis given the dynamic and unpredictable nature of highly virtualized environments; we term this as the Magic 8-Ball approach.

Also consider, when evaluating capacity planning and trending tools, the sample set that feeds your analysis – is it complete enough to make real decisions? For example, how can real determinations of capacity needs be met through a 30 day evaluation? Capacity planning can sometimes be more than an art than science; therefore the art is to make decisions guided through accurate and complete data. It’s similar to when you first install software onto your PC, if you were to take a 30 day snapshot from no software loaded to all the applications you loaded in the first 30 days – any trending and predictive analysis software would have to assume that

Open for discussion!

Anonymous