Turbonomic Review
It does things way beyond DRS and Storage DRS. Our past monitoring solutions just overwhelmed us with data.


Valuable Features

The Automation and Planning capabilities are unique, innovative, and game-changing for us.

Improvements to My Organization

We are able to maintain a much higher level of performance and efficiency in our virtual environment with the same or reduced levels of staffing.

Room for Improvement

We would like to see VMTurbo develop more application-specific modules that give us additional visibility into the end-user experience.

Use of Solution

18 Months.

Deployment Issues

No issues with deployment.

Stability Issues

No issues with stability.

Scalability Issues

No problems with scalability.

Customer Service and Technical Support

Customer Service:

Excellent.

Technical Support:

Excellent.

Previous Solutions

We used monitoring tools like Solarwinds Orion and vFoglight. They're fine products but they don't really compete with VMTurbo in that they are not actually performing command and control operations. VMTurbo actually does things way beyond DRS and Storage DRS. Our monitoring solutions just overwhelmed us with data and there's a big difference between knowing a problem exists and proactively preventing or fixing one without human involvement.

Initial Setup

We simply installed their VM and turned it on. Thirty minutes later we made the decision to go live with virtual machine automation and have not looked back.

Implementation Team

No, we did it ourselves with relatively minimal support from VMTurbo.

ROI

I would say we recouped our investment within two months. We were purchasing an average of six or seven hosts per year (at $20K-30K each). Based on the increased awareness of our infrastructure utilization rate, and VMTurbo-driven performance efficiencies, we eliminated the need to buy and deploy any new servers for 18 months.

Pricing, Setup Cost and Licensing

I think our startup cost was around $49K (plus another $40K or so for the VNX storage module eight months later). The day-to-day cost is about $50/day.

Other Advice

Just load the demo. It's completely non-invasive and you will probably be blown away by how much over-provisioning you have been doing.

Disclosure: IT Central Station contacted the reviewer to collect the review and to validate authenticity. The reviewer was referred by the vendor, but the review is not subject to editing or approval by the vendor.
2 visitors found this review helpful

3 Comments

it_user128907Real User

Hello George,

Thanks for providing good write up, I see a lot of VMWARE admins over provision resources in fear of spike demand to support early production hour. I would like to know how Vm turbo manages resource scalability to meet high compute, IO and memory demand for early 3 hours of business. Are there executive report that shows ROI after we implement VmTurbo? Looking forward for a demo.

Thx,
Kalpesh Surjiani

04 April 15
it_user215703Real User

Hi Kalpesh. Sorry for the delayed response.

I think it is safe to say that you are always going to have to provision and maintain enough physical and virtual resources to meet the high-demand periods. The nice thing about VMTurbo is that it will tell you what those requirements are based on trended metrics. If an application requires 4 procs and 8GB of RAM for three hours every day, VMTurbo is not going to recommend that the server configuration be modified on a daily basis to avoid over-provisioning for 21 hours. What it will do is help ensure that you don’t provision 8 procs and 10GB of RAM when you don’t need it. By the same token, if trending algorithms determine that you need to provision more or less resources, VMTurbo will make that recommendation to you (or actually reconfigure the resources if you have the appropriate automation capabilities enabled).

Since the loads on VMs, hosts, and storage arrays are constantly changing, resource starvation and associated latencies can come about at any time. VMTurbo, using its supply and demand market model is constantly looking for the cheapest (i.e. least utilized) resources for its “customers”, the VMs. As a result, it will make (or recommend) vmotion and/or storage vmotion changes in order to maximize performance and efficiency across clusters and virtual datacenters. VMTurbo actions are based on the pursuit of an always-changing desired state, not thresholds. According, it is always calculating the consequences of potential changes and evaluating the results of actual changes to ensure that there are minimal unintended consequences associated with relocations or any other configuration changes.

So, to reiterate, VMTurbo is unlikely to recommend daily server configuration changes to meet high demand periods, it will most certainly move VMs to optimal locations based on near-real-time resource supply and demand. If that is not effective enough, it will recommend adding server resources or the provisioning of addition servers, hosts, or storage. As an added benefit, all this can be modeled in the planning module (which is way cool!). I should also mention that in a number of cases, we improved application performance by reducing the number of processors in servers (usually dictated by app vendors) based on VMTurbo’s detection of ready-queue congestion.
Finally, I don’t know of any executive reports that specifically show ROI but when you run the default “plan” in the Plan Module, VMTurbo will give you a super-detailed step-by-step work-list showing every recommendation needed to get you to your initial “desired state”. For us, that plan included shutting down six hosts. We didn’t actually do that, but VMTurbo gave us the confidence to postpone our routine, annual purchase of five or six new hosts previously deemed necessary to deal with any performance issues (often brought to our attention by dissatisfied end-users). So our ROI was measured in delayed purchase of expensive hosts and by incredibly improved performance in our virtual environment. This technology is really worth a closer look!

I hope this helps you. I think it's fairly accurate as far as the product capabilities and characteristics are concerned.

--George

10 April 15
it_user128907Real User

Thanks George for detailed explanation on technical aspect and ROI, I recently came to know from my sources that Hyper-V next version would support auto scale of CPU and memory resources based on demand, we can call it as auto fat and auto slim :). I see great value in concept of HV Orchestration and how efficiently energy can be saved & utilised specifically in large scale DC. Eventually I should quote "Energy saved is money saved"

15 April 15
Guest
Why do you like it?

Sign Up with Email