For more than a generation, IT professionals have talked about the impending death of big iron mainframe computers. Twenty-five years ago, I engaged in debates at the Tier-1, multi-line carrier I worked for about what our future strategy should be. My comment at the time was that other types of computing would usurp mainframe domain space; I expected they would act more like “really large servers” than anything else. The resulting derision was memorable and, even now, I won’t proclaim that I was gifted with the clairvoyance to see that cloud computing would rise to prominence. Nevertheless, recent conversations with other carriers have rekindled that dialogue in a new context: the rapid evolution of computing power.
Legacy mainframe computer systems that depend on things like COBOL, Assembler, and VSAM files will never be sexy or cool again. With that out of the way, what do we see in the marketplace that carrier IT organizations should consider?
Reframing the Mainframe
One carrier recently stated that its goal was “to get everything off the mainframe and move all workloads to the cloud.” It is an interesting goal from a technology standpoint but moving workloads from one environment to another can produce surprising, unintended consequences, including cost increases.
Modern systems (with their agility and real-time access to information) running in modern cloud-based environments produce a wide array of significant business benefits under the right circumstances. The key is to understand what those circumstances are. Having an array of tools to choose from is better than becoming an expert on only one and then presuming that it applies to every use case.
Mainframe computers are still good at running legacy systems, many of which carry large, nightly workloads, running code that is difficult to manage but cheap to run on a unit basis. This is not a suitable environment for new products or distribution channel support. On the other hand, just moving an old system into a new space can create technical, operational, legal, and financial issues that require careful consideration.
Datacenters as a Service
Novarica has seen an increase of carriers working to “get out of the data center business.” This tactic, which presumes that computing power can be a commoditized service, becomes agnostic to what type of service is provided as long as it is appropriate to the job at hand. Some carriers use a cloud provider (e.g., AWS, MS Azure) as the target environment. An analog to this is transitioning big iron mainframe processing to a service provider who can run systems at a larger scale than individual carriers can, providing leverage around licensing and human capital. Both instances fundamentally change how IT organizations think about delivering capabilities to business partners.
These tactics help focus organizational intentions on financial and accounting related issues rather than on “technology for the sake of technology.” Moving to cloud-based solutions forces dialogue about how the organization will move from a CapEx world to an OpEx one. This can benefit organizations wrestling with the ever-shortening useful life of technology investments. Moving to cloud-based solutions requires CIO/CFO collaboration since it produces some fundamental changes in how organizations manage project financials by, for example, moving away from what may by today’s standards seem like unreasonably long depreciation cycles.
Taking big iron and moving it into someone else’s data center can have a similar impact. Pricing becomes clearer, and depreciation cycles (which may mask how benefits emerging from new investment business cases) become more transparent.
What we see now is a fascinating evolution in the insurance business and the underlying technology that supports it. Having the right tools for the right workloads should be a guiding principle for the technology strategies that CEOs and their teams delivering on today. Novarica has an expertise that may be timely for companies that are considering how to build those strategies.
Add new comment