The English village of Bucklebury, which lies about two miles from my home, has achieved instant fame as the family home of Kate Middleton, Prince William’s new wife and the U.K.’s future queen. In the weeks leading up to the royal wedding, the village was besieged by journalists and paparazzi of every nationality, as well as curious sightseers and coach trippers, all anxious to catch a glimpse of the duck ponds, quaint shops, and thatched roofs that reportedly characterized the princess’s idyllic existence. Bucklebury provided all the ingredients for a Disney-esque, fairy-tale romance, and the media and PR specialists molded them into a tasty dish for worldwide distribution.
Now, I don’t wish to shatter any illusions. I loved the wedding celebrations as much as the next guy and Bucklebury is a very pleasant village. But I doubt that Kate Middleton used to dance around the Maypole on her way to school in quite the way dramatizations might suggest. This is an example of the power of the media to create impressions—impressions that last and become ingrained in the subconscious in a way that becomes difficult to dispute.
The image of the mainframe has been similarly (but not so positively) molded by the industry gurus and wordsmiths, and the many negative myths surrounding the platform continue to hamper those who see real business advantage in running new System z workloads. During a “scheduled chat” at CA Technologies’ recent May Mainframe Madness virtual conference, which I was invited to attend with my associate Trevor Eddolls, co-author of the Arcati Mainframe Yearbook, I was reminded just how many organizations face this challenge every day. Users taking part in the discussion recounted their experiences of running 100-plus Linux on System z images on a single box, consolidating numerous distributed processors with associated manageability, power and space benefits—not to mention dramatic cost improvements. But winning the confidence of managers and budget holders, persuading them that System z is the way to go, remains a slow, painstaking process.
One user said she had been given the go-ahead to port just one application to Linux on System z. On the strength of this very successful implementation, she was gradually winning the support of managers who had been brought up on Windows and UNIX, dispelling their strongly held belief that the System z is an excessively expensive and complex platform.
All too rarely, the business media decides to put a positive spin on migrations to the mainframe (as in the case of Allianz in Australia, which moved its whole server farm to Linux on System z in a weekend last year with negligible outage). But most commentators are happier to reinforce the legacy myth that has endured for years—and in so doing, they put obstacles in the way of those who are making a concerted effort to bring imaginative cost and efficiency savings to their businesses.
A great CA Technologies online conference, by the way!
Around the Vendors
American Express must be feeling like a million dollars, and that’s exactly what they’ve won as a grand prize in Compuware’s Mainframe Cost Savings Program. Participants in the program were offered a free cost-savings assessment, which identified current opportunities and provided a roadmap for future operational cost-savings for the mainframe. AmEx came out on top as the company that had achieved the greatest savings with the help of Compuware tools. “The more efficiently we operate, the more savings we have to invest in growth opportunities,” says the company’s senior vice president Matthew Robinson.
Canadian mainframe performance optimization specialist DataKinetics has released a Java to DB2 Optimizer. The product, originally developed for a bank that was experiencing unexpected increases in data access request time, promises to offer a very significant improvement in Java application performance within a DB2 environment.
Pioneering z/OS network management company William Data Systems has introduced a “peek” function with Version 4.7 of its ZEN TRACE & SOLVE product. The function provides in-flight viewing of active IP traces. This means that it’s now possible to view an active trace while it’s still running, tracking its progress in-flight, and then stopping it when the condition being sought in the trace is detected. This, says the vendor, can save considerable CPU costs and user time.