Where applications and infrastructure meet
There are a lot of people out there talking about the importance of applications in todays data center. There’s a lot of smart things being said, but to use a phrase inspired from my ham radio days, the signal to noise ratio is pretty poor.
Applications are, of course, the key component in todays enterprise: they are the portal for business delivery. Furthermore, as infrastructure hardware continues to commoditize, its just not a sexy topic to talk about anymore. But the fact of the matter is you cant have one without the other. Applications need hardware to run on and businesses need application to run. Looking at one without understanding the other is short sighted.
Understanding the complex relationships that exist between applications and the shared hardware infrastructure that supports them has become an important issue for IT organizations. One trend that is making these relationships even more difficult to grasp is virtualization. Virtualization is hot, but it vastly complicates the process of getting a complete picture of these relationships.
So let’s explore virtualization in more detail for a moment. Everybody seems to be talking about server, network or storage virtualization, but in reality, virtualization does two things. First it turns physical resources into services. With virtualization, CPUs, networks, and data storage all become fungible. This allows these physical resources to be delivered as services. Over the long term, those services should be able to be delivered on-demand, changing in real time as business requirements change.
Secondly, virtualization adds lots of complexity to the IT infrastructure. It has, in fact, become too complex for humans to understand how multiple applications interact with a shared, virtualized infrastructure. Just imagine how difficult it is today to understand how one application sitting on a virtual server–connected to a virtual network, talking with virtualized storage–interacts with another applications being serviced by the same infrastructure. Now ask yourself, which application is accessing which drive at what time? This is a fact of life for today’s IT organizations, and if you’re not able to answer that question when trouble hits, when you’re planning your capacity growth or when your committing to service levels, you’ll be in trouble.
This is one of the many challenges that I set out to solve when I founded Akorri. We’ve developed an application that you’ll be reading about shortly that helps IT organizations understand the complex relationships that exist between the cross-domain technology silos in todays data center–helping them balance performance, availability and utilization/cost for instance.
Using complex analytic and modeling algorithms, Akorri can help you identify and troubleshoot performance and availability related issues, provide optimization recommendations and allow effective, accurate planning.
In future posts Ill discuss how this technology works, the importance of analytics and modeling, and how customers are already using our technology to solve some important challenges.
