Since the early 1990s IT pundits have been declaring the mainframe dead. With almost every IT advance: x86 virtualization, multi-core x86 processors, cloud computing some IT analyst announces the end of the mainframe. Those declarations have slowed since IBM introduced the hybrid zEnterprise 196 in July 2010 and the mainframe experienced a series of impressive quarterly sales gains.
That’s what makes this latest mainframe obituary so surprising. A white paper from Micro Focus reports that “CIOs are increasingly questioning whether the mainframe will continue to be a strategic platform in the future. Written by Standish Group and based on its CIO survey that found 70% of respondents said the mainframe provides a central, strategic role currently. However, none of the CIOs consider the mainframe as a strategic platform in 5-10 years time.
None? Zero? Nada? That’s pretty astonishing. So, what are the CIOs’ complaints? The study isn’t exactly specific, but it seems to do with the cloud.
As Standish puts it: looming large on the CIO agenda is cloud. Cloud creates both challenge and opportunity for CIOs today. The opportunity lies in driving towards more flexible, cost-effective service provision for the business, enabling in-house IT resources to focus on much more strategic initiatives. At the same time, CIOs are managing a host of current technologies and applications, some applications duplicating others, some legacy applications for which there appears no easy modernization solution and entrenched solutions and applications that provide no clear journey to cloud-based services.
Whoa, let’s parse that sentence. As Standish sees it, CIOs will look to the cloud for flexibility and cost-effective service provisioning that frees IT to focus strategically. Based on that, you could just as easily build the case for the zEnterprise, starting with the entry z114 and the Unified Resource Manager.
But that’s not really the issue; IT modernization is. The researchers note that the need to address legacy mainframe applications effectively is a critical success factor. Furthermore Standish observes: these applications are still used in the organization today [which] emphasizes their business importance, and often there is a high level of intellectual capital embedded within these systems.
OK, so the real complaint is around leveraging legacy applications as valuable software assets. Fortunately CIOs can do this without undertaking a rip-and-replace of the mainframe. SOA is one place to start. SOA provides way to extract business logic from mainframe apps and use it as services. The mainframe does SOA very well. Independent Assessment, the publisher of BottomlineIT, has written a number of case studies on mainframe SOA. Check it out here and here.
Standish digs up a few other complaints about the mainframe, such as the shortage of mainframe skills and the high cost of mainframe computing. These are old complaints and much is being done to address them. With the z114 and the System z Solution Edition Programs IBM even is putting a dent in the cost-of-acquisition issue.
Then the paper offers this intriguing complaint: Being forced into a decision to move from unsupported mainframe environments to continue operations and meet new performance levels. Huh? If you are seeking to meet a variety of new operational and performance levels while efficiently managing and supporting it all the hybrid zEnterprise seems made to order with z/OS, z/VM, Linux on z, specialty engines, AIX on Power blades, and soon x86 on a blade. Standish seems oblivious to all the changes the mainframe has undergone since the introduction of the zEnterprise over a year ago.
This, however, is a Micro Focus paper so Standish isn’t interested in looking at how mainframe shops can leverage what IBM has been building into the mainframe and zEnterprise over the last few years. Yet, to position themselves for cloud computing, private clouds, and to meet the CIOs’ reported three 2014 top objectives—1) increasing enterprise growth, 2) improving operations, and 3) attracting and retaining new customers—the zEnterprise is exactly what they should be looking at. Instead, Standish recommends rehosting and migrating applications, and how best to do that is with, of course, Micro Focus.
Not every organization or workload should have a mainframe. Many don’t. Similarly, there are situations that can be best dealt with by migrating mainframe applications to a different platform, but cloud computing probably is not one of those because the mainframe can play very well in the cloud. It would have been nice if Standish had focused on those workloads and situations that make sense to rehost and migrate while at least acknowledging the new hybrid mainframe world.
Please note: DancingDinosaur will be unavailable next week and not able to moderate comments until 9/17.