Leprechauns, Unicorns and Storage IO?
by Jonathan Reeve August 13th, 2013
There are a lot of myths and legends when it comes to managing storage IO in virtualized environments (you were wondering where leprechauns and unicorns came in, right?). Let’s clear a few of those up and how VMTurbo can help.
Storage performance continues to be one of the major challenges to assuring performance in virtualized environments. VMTurbo takes a fundamentally different approach to assuring performance in virtualized environments. It solves what we call the “Intelligent Workload Management” problem or “how to assure workload performance while maximizing the efficiency of infrastructure.” In the context of storage, this means assuring storage performance to workloads while utilizing storage resources as efficiently as possible. These are conflicting goals, and many organizations resort to drastic over-provisioning—and the resulting costs of storage resources—in an attempt to assure performance. VMTurbo measures storage IO, latency and storage space, and factors these demands into it’s IT marketplace-based model to ensure that workloads receive adequate storage performance at all times while efficiently leveraging resources from host and datastore providers.
Planning and Onboarding
VMTurbo takes into account both storage IO and network IO when planning across multiple scenarios, or helping to optimally onboard new workloads. The key to providing rapid and accurate plans is in solving the Intelligent Workload Management problem not just in real time, but also in the future. Scenarios such as adding or removing additional virtual machines or hosts require a global view of the infrastructure. That is, understanding that placing a new virtual machine on a host can have “knock on” affects on other virtual machines and ripple effects through the infrastructure.
The Storage Beneath the Virtualization Layer
With the recent 4.0 release of VMTurbo Operations Manager, control of storage IO has been extended into the storage beneath the virtualization layer. This extension maps from “VM to spindle” to provide real-time control and optimization of both domains to obtain desired storage efficiency gains and workload storage performance assurance.
VMTurbo discovers the types and IO capacities of the underlying disks used by the storage arrays under management, allowing for the IO capacity of the underlying storage entities to be calculated (volumes, aggregates). As IOPs usage driven by workloads becomes high (and therefore more expensive in the VMTurbo marketplace), decisions regarding resource allocation (resize/create volumes and/or aggregates) and/or workload placement (to satisfy workload demands to reach an optimal equilibrium) may be required.
As you can see, storage IO is a very important consideration in VMTurbo’s approach to deploying, operating and planning across virtualization and storage domains. And, unlike leprechauns and unicorns, it is available right now.