Cisco & Turbonomic - Success at Scale

Video created by angelo.luciani on Jul 13, 2017

    Mike Myers: Cisco IT supports over 30 millions watts of raised floor that's literally billions of dollars of hardware spread across our lab environments and our data center environments. It's all Cisco hardware. So, we're using the exact same things that we build. We have amazing scale to demonstrate what some of these products are really capable of doing and when they fit into the ecosystem so well, it just becomes a natural extension of that Cisco-powered data center.

     

    Every good data center starts off with Cisco hardware. Once you have a fully programmable infrastructure, then you can start to layer other capabilities on top of it. Some of those capabilities come from our own teams, like Tetration. The next component that's going to be part of that stack is going to be the Turbonomic suite.

     

    Chris Heinbokel: Turbonomic enables elasticity for us by continuously sweeping the environment to identify what clusters are running hot, what clusters are running cold, what VMs are actually using all of what they have today and may need more. And simultaneously what VMs are sitting there idol but have 28, 32 gigabytes of memory allocated to them, sitting unused.

     

    You really see the enablement of elasticity, not just necessarily resizing your VM, whether you're resizing it smaller or larger, but actually increasing the performance of your VM. We saw this in our RTP1 data center. We saw in RTP1, that enabling elasticity across about half of our overall infrastructure and downsizing about a quarter of it, actually reduced our overall resource contention in the data center by 80%. 

     

    Mike Myers: We put all these tools in place. We understand what we have and how we can scale it and how we can grow it. Then all of a sudden, we don't need to spend $17 million. But more importantly, we don't even need to spend the $3 million a year to rent RTB7. So the first year, it was going to be the $17 million savings and then subsequently in propertuity $3 million a year, actually $2.8 million a year. But it saved us 54 terabytes of memory.It saved us, I want to say, 4,800 virtual CPU cores. So, we could add, after that reboot, another 2,200 servers for free to the environment.

     

    Chris Heinbokel: We'll be able to do more with our existing infrastructure. We'll be able to make the most out of our data centers and your VMs will run better. What's not to love about that?

     

    Mike Myers: So, when you look at a Cisco-powered data center, the Turbonomic suite with its dynamic model, is really the secret sauce for how you get everything out of that data center that you can. Okay, I've decided that I want to move into the public cloud. I know that it's going to cost me a fair amount of money to move there. I know that it's going to be difficult for me to move back. So, let me optimize my data center. If you don't optimize your workloads before you put them into the cloud, you're going to spend a ridiculous fortune on running those workloads in the cloud. But, when you really look at your data center constellation, if you've optimized it appropriately, it's going to be able to compete with the public cloud, probably for nine out of ten use cases. You're just really going to get your mileage out of running your environments very, very efficiently. Hopefully, those decisions become more and more automated over time so that we really get to the autonomic capabilities that we would want coming from the environment. I throw a switch, it works, it works perfectly every time. I meet the SLAs, I realize the revenue and we always are really optimizing our flexible assets in the data center to their potential.

     

    Chris Heinbokel: We couldn't have done it without Turbonomic.

     

    Consider sharing on social media and continue the conversation in our product forums.