Migrating applications from IBM IMS mainframes over to newer open systems requires mainframe compatible transaction capabilities. Developers can continue to program within familiar IBM IMS constructs or move to newer frameworks as they see fit.

According to Clerity’s press release, mainframe migration has already saved the IT Department of the Ministry of Justice (ITIJ) in Portugal an estimated 95% on annual operating expenses. ITIJ had already moved many of their operations over to open systems servers, but two mission-critical applications still ran on the IBM IMS style mainframe. These two applications alone were costing the ITIJ approximately 1.45 million Euros per year. By decommissioning the mainframe, ITIJ dramatically lowered costs while simultaneously increasing systems flexibility.

In his CMG’09 talk, “A Mainframe Guy Discovers SOA and Cloud Computing,” IBM Manager of IT Training Glenn Anderson looks at the next leap for mainframes now that SOA has become a reality. Anderson has been an IBM mainframe guy all his career. He said SOA with cloud computing is one of the biggest evolutions since the client-server model hit the scene.

“I think what’s more interesting to us mainframe guys is this notion of the private cloud,” said Anderson.
“Which means changing the internals of your IT organization to model the public cloud.”

Over the past few years, IT teams have been thinking more and more in terms of service orientation. But when it comes to the public cloud, most enterprises are concerned about the security risks of storing sensitive data in external, multi-tenant environments.

But since so many data centers are now service enabled, Anderson said, mirroring some of the principles of cloud computing is a next logical step for many. Anderson said mainframe shops could benefit from being able to quickly scale virtual system images up or down to deliver mainframes like a service.

“If a mainframe data center wants to be flexible and responsive to its business,” Anderson said, “you’ve got to be focused on delivering service to the business in a way that makes it easy and productive for the end users.”

Many mainframes today already are service enabled or working with a SOA stack of some kind. This, Anderson said, was the major re-architecting that had to happen to turn the data center into a private cloud environment. The next step is a matter of getting a bunch of machine images to live inside single partitions.

Having a number of virtual images sharing a common partition makes them easier to manage, Anderson said, and can be a more efficient use of space.

As the Computer Measurement Group’s (CMG) 35th annual CMG Conference kicks off in Dallas, Texas next week, cloud computing will be the center of attention. A number of sessions will cover related topics that focus on data and infrastructure management in the cloud.

CMG, originally associated with mainframe technology, is a long-standing organization of data processing experts who work on the measurement and management of computer systems. Over the years, the group has worked to expand the well-honed performance- and capacity-oriented tools and disciplines for use in server-centric computing.

As cloud computing vies for attention, the CMG has undertaken new efforts to explore related performance issues. Various viewers have suggested new takes on mainframe methods will come to the fore in the cloud.

Industry analyst Amy Wohl, president of Wohl Associates, will deliver the keynote CMG address titled “The Impact of Software as a Service (SaaS) on the Enterprise Data Center.” Following this, Jeffrey Nick, Senior VP and CTO at EMC Corp will lead the plenary session on using virtualized data centers in the cloud.

Other topics of discussion include business service governance, infrastructure on demand, capacity planning, data crunching in the cloud and–no surprise, here–performance and scalability.