The evolution of operational efficiency

by Larry Spangler (Red Hat)

Lately, I've been seeing and hearing a lot of buzz about "operational efficiency." As some see it, Operational Efficiency is basically the idea of doing more with less--if you can define and follow processes you can achieve repeatable outcomes with reduced error. Automate that, and you have a means to extend the reach of the individual IT operator while decreasing the effort and time required to build systems. It's a straightforward value proposition that Red Hat has been touting and delivering for years with standardized operating environments (SOEs) and management tools like Red Hat Network Satellite and JBoss Operations Network.

But there's evolution afoot here from the classic "operational" sense to one that is more expansive and higher purposed. The basics of SOE and management tools are now being used not only to define and develop repeatable infrastructure, they're being leveraged with other tools like virtualization, IaaS, and PaaS to deliver on-demand capabilities. The key being that the focus is shifting from how to get the most out of your resource investment, to how to effectively and efficiently instantiate, use, and release systems for true on-demand capabilities.

One of my earliest efforts in the industry was focused around assessing and improving capability maturity. Back then, the thought of variable-driven build scripts to configure test and production servers was fairly new and by no means pervasive. We spent time understanding how our development processes and products were delivered into the lab for testing and into the production environment. Writing a one-time script for a given build was easy enough, but it wasn't until we began treating that script as a managed element of the product with its own development and test plan that we started to see a true evolution of the system with direct impact on delivery quality.

Now, years later, we're seeing the same theme unfold yet again. Virtualization has brought us to the point where we can do more with less, but there are still many organizations where the practice of developing and implementing builds is more art form than science. Compound that with the fact that there are few tools available to manage build specification and implementation and you have an opportunity for improvement and evolution.

This is where the new generation of technology is taking us. We're seeing tools and concepts hit the market now that have the ability to capture build specifications that go beyond individual systems to encompass multi-tennant systems, Infrastructure as a Service, and Platform as a Service. To me, this means we're less focused on the typical assembly line process for building capabilities and more focused on employing programmable assembly robots that can do different tasks at different times of the day or month as the end product or capability demands. The beauty is that the manufacturing line can now change mid-stream from one capability to the next, or can actually deliver multiple capabilities simultaneously.

We've already helped clients break free of stove-piped dev/test hardware boundaries, leveraging self-service, on-demand, mini-platforms to meet the ebb and flow of their development teams' needs. We've already begun consolidating control and management aspects for these on-demand capabilities across multi-vendor resource pools. Now we're seeing an opportunity around modernizing IT from a bespoke, less-than-dynamic machine to a multi-purpose, vibrant organism that adapts to the shifting demands of the business. My suggestion--if you haven't already begun your own operational efficiency evolution, you should, because from SOE to infrastructure on-demand to dynamic platform environments and beyond, it's a fascinating time with much more excitement and opportunity ahead.

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.