In May 2000 IBM announced Linux as a fully supported mainframe option. Now, ten years later nearly half (over 3000) of the 6300 System z applications are Linux. Here is a list of about 60 of the latest new or upgraded Linux on z applications from ISVs.
Linux on z changed mainframe computing in major ways. I’d even go so far as to say Linux on z along with Java and SOA saved the mainframe from a slow, disheartening fade into irrelevance. But more than anything else it was Linux on z.
By May 2000, IBM already owned the mainframe market. Mainframes provided the core transaction processing for large enterprises that needed extremely secure and reliable processing of high volumes of transactions. Every organization needing that capability already had a mainframe. If they didn’t by then they probably never would. And even shops that had a mainframe were running other workloads on various distributed systems, sometimes big UNIX boxes from IBM, HP, and Sun or, increasingly, on x86-based platforms running Linux.
Where IBM sells new mainframes to first-time customers today is mainly in the emerging world of Southeast Asia, China, Korea, and now Africa. That was the case with BC Card, a Korean payment processor. See my case study here.
Linux on z expanded the range of workloads companies could run on their mainframe. Now IBM could go back to its mainframe customers and make a compelling case for moving some of those distributed applications to Linux on z. By discounting the price of the IFL and playing with the software licensing terms so companies could avoid incurring high software costs, IBM could come up with a compelling price. It worked; companies slowly started consolidating distributed workloads on their existing System z. Salt River Project is a good example.
z/VM only made the IBM Linux on z pitch better. Over the last 10 years, x86-based virtualization became hot. Well, z/VM offers a more robust, industrial-strength virtualization engine. With z/VM companies could run a hundred or more Linux images where x86-platforms could handle a few dozen at best. This boosted the already attractive Linux on z ROI story.
At the same time, SOA was experiencing a wave of adoption. Organizations began turning to SOA to let different applications and platforms work together. Now systems could access functionality on other systems and exchange data as standards-based services. SOA turned the mainframe, long a rich trove of proven business functionality and critical data, into an even more valuable asset. Ball State University, for example, used SOA with the System z and CICS to let its students access their critical class schedule over the Internet, even directly from Facebook. Check out the case study here.
So Linux has proven it can expand the workloads on the mainframe. However, it alone is not enough to attract enough new mainframe customers. To do that, IBM needs to substantially lower the total cost of mainframe acquisition. This is where the System z Solution Edition program comes in, which bundles different combinations of software and middleware with a deeply discounted System z and 3-5 years of maintenance. IBM recently stated it will continue the System z Solution Edition program with the new System z coming later this year.