Gino Pokluda had a problem: The database system at Presbyterian Health Plan in Albuquerque, N.M., where Pokluda serves as manager of service improvement and innovation, was becoming increasingly expensive and unwieldy, requiring about 80TB of storage for 13 database environments. To gain control, Pokluda implemented Delphix software to enable agile data management and eliminate redundant infrastructure. The 2012 project sliced his storage needs to 35TB, even though his team now maintains 23 environments. Here Pokluda, who manages all production, test and development environments for the company, discusses the database system overhaul and shares other IT management insights.
What are your key responsibilities? To look at those things we do now in IT and to get a culture of innovation to take hold here. That's something that hasn't existed in the past. In addition, I'm charged with implementing ITIL best practices throughout IT.
How do you define innovation? I can tell you the things it's not. It's not process improvement. Process improvement is where innovation gets hijacked, and it always revolves around ROI. Innovation, in my mind, is, "What is the job to be done and what is the best possible way to do that?" Innovation doesn't always involve technology. It could be just looking at something differently.
How do you cultivate innovation? The best innovation comes from the bottom up. You get those who are doing the work, and you give them the opportunity to come up with new ideas. We have 100 people in IT, so we have 100 innovators, and I submit that everyone has at some point innovated and they just don't realize it.
Was the Delphix project innovation or process improvement? It was definitely innovation. We were a mainframe shop that got thrust into the world of diversified processing. That made life a whole lot simpler. But in 2005, our payer system [vendor] decided that they were not going to provide any more upgrades. So we had to enter the world of Windows servers and a diversified infrastructure running Facets [a healthcare payer system]. In 2005 we built this Windows architecture running this product and, unfortunately, we had a flawed architecture. As a result, this became a legacy problem we perpetuated until about two years ago when we realized we couldn't keep up with the business demand for databases. That prompted us to look at how we maintain storage and utilize our storage.
What finally prompted the overhaul? Our biggest problem was volume: We had a 1.5-terabyte database for production and a 1.5-terabyte database for development. Multiply that by three or four, and then a couple of configuration environments and testing environments and training environments, and 33% growth each year. All of a sudden, the cost of storage for nonproduction environments is rising exponentially. When we got to 13 environments, we said something's got to change. Also, we could not meet the needs of the business. If they had a large project or a large push for a regulatory requirement and they had to test it, we just couldn't do it.
What sold you on this technology? The virtual databases the product provided acted just like the [original] databases. They were every bit a database except when you looked at the footprint.
How did you devise the solution? I did what everyone else does: I Googled. I was fishing for clues; I was fishing for something out there to see what everyone else was doing, and that's where I came across a white paper written by Delphix for Boeing about how Boeing was experiencing these same issues in their credit union and how they used virtualized databases to solve their problem. We were already well into our journey into VMware, virtualizing our servers. And then you realize that if you can virtualize databases, you can virtualize your entire stack. And when you combine those, virtual servers and virtual databases, you can clone entire environments very easily. I realized that was the way we should go.
What was the biggest challenge? Actually persuading corporate leadership to go this route; convincing them that this new technology was going to be beneficial to them. We convinced them through a number of presentations, and we sold it to them by saying we'll try it for a year and see how it works.
Besides lower storage requirements, what other benefits did you gain with this project? It used to take 50-some-odd days to develop an insurance product, to get that through configuration, development and testing. That time frame has shrunk. Now we can get a product to the customer in about 23 days.
What was the biggest lesson learned from that project? It doesn't have anything do to with the product, but I've learned that IT shops, when they're deeply embedded with their customers, the relationship is more of an ecosystem, because something we do here can affect something down the line or up the line. So the relationship between IT and the customer is approaching a symbiotic relationship rather than a collaboration. I've seen how much this product has enabled us to help the business.