June 26, 2014 by Bill Johnson
The world doesn’t run on Intel INTC -0.74% alone. At first glance, it may seem like that, but experience has shown us that things are never as simple as they seem. As an IT availability company managing and hosting services for more than 7,000 customers, we have unique insight into the realities of an IT environment. To prove our point, we recently looked at the requirements of more than 1,800 of our recovery-as-a-service customers to discover just how complex the business IT world is. The results were a “hot mess,” to say the least.
While it’s true that nearly 9 out of 10 customers (87 percent) depend on the Intel x86 platform for critical business functions, only about a third of them (35 percent) run on Intel platforms alone (source: Sungard Availability Services data, June 2014). We also learned that the overwhelming majority — 65 percent — run on hybrid systems that integrate Intel with an alphabet soup of other platforms (AIX, IBM IBM -0.48%, SUN, VAX, UNISYS, DEC, and more). In other words, it’s a hot mess.
What does this mean for IT leaders? Well, it means that successfully accomplishing IT disaster recovery and meeting the goals of ongoing resiliency initiatives are much harder than they realize.
What’s So Hard About Recovering Hybrid Environments?
In order to successfully recover your production environment, you have to replicate the precise mix of servers, storage, operating systems, hypervisors, networks and software versions in your offsite recovery environment. Without such precise orchestration, recovery will fail.
If your environment is hybrid, then you need to replicate more kinds of technologies and plan for more nitpicky considerations, like the need for purpose-built machines and specific configurations or versions. More variables in the environment mean higher complexity and an increased likelihood of delays. And as they say, time is money.
Let me give you an example: a regional insurance client of ours grew their IT infrastructure over 25 years with multiple acquisitions. By the time we got to them, they had 92 Intel machines, 89 AIX servers, and 250+ virtual machines supporting a mix of applications.
Their business processes (processing a claim, sending adjusters to an accident, etc.) all worked in concert, so they needed consistent backup and recovery times across ALL of their applications and infrastructure. The problem was, these business processes were supported by a mix of technologies, all with different recovery times. Some were fast – a few hours – while some were slow – four days. What made things worse was that some of their business-critical applications depended on those technologies that took longer to recover.
The total recovery time for the entire system to be fully functional again, i.e., adjusters could start adjusting and customer service could start answering questions, was therefore much too long. It’s like they say: “You are only as strong as your weakest link.”