Back to Blog

Andrew Mallaband

A “Lightweight” Approach To Ensure Performance and Efficiency?

VMTurbo is just a lightweight scheduler.”


That’s the perception that a prospective customer I visited this week had about VMTurbo Operations Manager. He shared that sentiment at the top of our discussion and was convinced that we could not help him based on his understanding. He told to me, “VMTurbo Operations Manager shuts down low priority applications without regard to application function or need, requires lots of effort to set up, and conflicts with VMware DRS.”

Being extremely puzzled by this, I asked him, “are you thinking about another solution?” I pressed on and shared the results our customers are achieving with our solution.  His perception of VMTurbo completely changed.

To me, the definition of “light” means “having less substance or weight.” On one hand, you can describe VMTurbo as “lightweight” because it only uses a small footprint of compute and storage to operate, it deploys in minutes, and delivers immediate value without the need for professional service or the ongoing overhead associated with administration. On the other hand, VMTurbo addresses a “heavyweight” problem facing many service providers and enterprises:  “how to ensure workload performance while maximizing the efficiency of their infrastructure.” At VMTurbo, we describe this as the “Intelligent Workload Management Problem.”

Today, many organizations are trying to solve this problem using solutions like DRS, which are focused on optimizing a type of resource without regard for the broader picture of resource and business constraints in the data center. These are complemented with monitoring solutions and IT Operations staff to identify the performance issues solutions like DRS do not address, as well as IT Operations staff to resolve them. The reality is that this is a reactive approach that does not scale and infrastructure and operational costs escalate dramatically, reducing the overall competitiveness of internal IT.

Fighting robot figurines

VMTurbo is tackling the “Intelligent Workload Management” problem with a holistic software solution. Solving this problem requires that you look at a broad set of resources (CPU, CPU Ready Queue, Memory, IO, Network IO, vMem, vCPU, Virtual Data Centers, Disk Space, and more) and constraints  across the IT stack to determine a broad set of actions (workload placement and sizing of physical and virtual resource allocation) on an ongoing basis to prevent performance bottlenecks while maximizing resource utilization.

VMTurbo does this very elegantly without conflicting with DRS. In fact, it delivers incremental improvements in memory and CPU performance and efficiency than can be delivered with DRS, reducing the % deviation in utilization between hosts in a cluster. VMTurbo also automatically imports Affinity and Anti-Affinity rules configured in DRS to ensure that pre-defined placement policies are respected in the operational environment. It is very common that some of these policies were originally put in place to control the behavior of DRS and many customers find that they no longer need these once they run VMTurbo—simplifying the ongoing administration of their environment.

This is clearly illustrated in a real-world environment through the following report which shows the memory convergence in a six-node cluster that is running DRS.

VMTurbo report showing memory convergence in six-node cluster running DRS

If you are facing challenges assuring workload performance and infrastructure efficiency of your virtualized data center, VMTurbo has a “lightweight” approach to a “heavyweight” problem. Try it for yourself.