NYC Summit Exhibitor Focus: Turbonomic

In anticipation of our New York City Summit Event on October 19th, we’re highlighting some of the great partners who will be in attendance in this Exhibitor Series. We’re excited to share our latest Q&A with Charles Crouchman, Chief Technology Officer of Turbonomic. 

AppD: What was the impetus behind Turbonomic being founded?

CC: Turbonomic was founded on the premise that software can manage virtualized IT systems better than human beings to assure performance while maximizing efficiency. Turbonomic’s algorithm applies microeconomic theory and the principles of supply and demand to resource management in the data center and cloud stack.

This concept of applying economics to shared compute resources was originally discussed in a series of papers by co-founder, Yechiam Yemini, in the 1980s. The fundamental concept is that workloads choose the infrastructure on which they run and consume only the resources they need to perform, in the same way that shoppers look for the best overall price for a basket of goods, buying only what they need, when they need it.

AppD: Why does overprovisioning in the cloud occur, and why is it such an issue?

CC: Overprovisioning in the cloud occurs for two primary reasons. First, virtually every workload residing in a private data center today is overprovisioned because there is no cost-efficiency penalty for doing so.

Second, cloud providers actually encourage enterprises to overprovision their workloads for performance reasons. Combined, migrated workloads are migrated into oversized templates and net-new public cloud workloads are oversized from inception. The problem lies in the fact that overprovisioned resources are now rented, not owned, and the cost overruns can be significant. For example, if a customer over-sizes just 100 workloads by one template in Amazon Web Services, it equates to $1.2M in annual unnecessary spend.

AppD: How does Turbonomic reduce 30% or so of typical cloud provider costs?

CC: Turbonomic continuously observes workload demand, also known as consumption, and matches that demand to the cheapest available supply. The platform understands the actual resource consumption of each of your public cloud instances. It then matches that consumption to the best available template—the one with just the right amount of compute and storage—resulting in specific actions to resize templates, scale applications, or shift storage tiers. These actions bring supply and demand into alignment.

Where does the 30% come from, you ask? Well this is typically the average amount of resource and dollars by which customers overprovision. With Turbonomic at work, that waste can be safely eliminated.

AppD: What makes Turbonomic so different to other cloud-related vendors?

CC: Turbonomic bears several key differentiators, but the standout point is that Turbonomic is the only real-time management platform that bridges on-premises and cloud infrastructure for performance, cost and compliance management. We integrate with multiple providers up and down the stack, are entirely API-driven, and leverage the aforementioned economic abstraction, which empowers our agnostic approach.

At the end of the day however, it is the fact that Turbonomic is the only platform on the market that can assure your workloads are running performant, at the lowest possible cost, within compliance, regardless of where they reside – all in real-time. Nobody else can offer that, or anything close to it today.

AppD: Tell us about your decision engine that helps determine where a workload should run and when.

CC: The fundamental idea is that the workloads choose the infrastructure on which they run. If you have ever purchased a stock, or bought something on eBay, or spent money on anything, the item you purchased and the amount you paid presumably represented your demand for that item at the time. Turbonomic works in much the same way. Our decision engine exposes the realm of possible residences for each workload… On-prem VMware, Hyper-V, Amazon Web Services, Microsoft Azure… All zone, regions and clusters.

It then prices access to those entities as a function of their real-time utilization and the real-time utilization of their constituent resources (thread pools, heap, database connections, network throughput, storage latency, etc.). This pricing mechanism is a virtual currency, not a real one, but the net result is that workloads shop for the best deal, based on their resource requirements at the time. When a workload identifies a need for X, Y and Z compute, storage and network resources, Turbonomic brokers the purchase, so to speak, and the action to migrate and/or scale the workload takes place.

AppD: How does Turbonomic work with Cisco and AppDynamics to benefit enterprises?

CC: Customers are recognizing that applications are the lifeblood of today’s business. With Cisco and AppDynamics, Turbonomic is bringing our core value props—performance, efficiency, and compliance—to Cisco’s intent-based data center and enabling IT to focus on applications. With these partnerships, every decision our platform makes is about matching real-time application demand to the underlying compute, storage, and network.

The added benefit for UCS customers is that the Turbonomic integration enables demand-based elasticity in the UCS environment, turning blades on or off based on the resource needs of the applications running on UCS. For AppDynamics customers, Turbonomic can discover the application topology and OS metrics through AppDynamics and maps it to the data center stack. Using that information, the decision engine ensures that application components and workloads across the stack get the resources they need when they need them. With Cisco and AppDynamics, we’re stitching application performance directly to self-managing, elastic infrastructure.

AppD: Why should delegates at the NYC Summit visit the Turbonomic booth?

CC: Do you want to see how self-managing infrastructure directly impacts application performance? Come to the booth. We’ll have a demo showing how Turbonomic pulls in application topology and metrics from AppDynamics and then uses that information to drive the right sizing, placement, and capacity decisions that improve performance.

Applications and infrastructure have traditionally been siloed. With this integration, the finger pointing and guessing games are over—this is a whole new ball game for IT—one that every CEO, CRO, CMO, and, yes, every CIO will care about.  Because, again, today’s organizations rely on applications and the business outcomes they drive. Only software can assure their performance in real time, all the time.

Register here to book your free place at the NYC Summit on October 19th and meet the Turbonomic team there.

Charles is Chief Technology Officer of Turbonomic. In this role he contributes to product strategy, evangelizes our technology, supports our strategic sales and business development efforts, and leads product management. Prior to joining Turbonomic, Charles held senior executive positions at several technology startups including Cirba, Mformation Technologies, Opalis Software, and Cybermation.