A year ago Mac Devine, IBM Distinguished Engineer, extolled the virtues of the mainframe as the foundation for a SaaS strategy. Since then a slew of z cloud initiatives by IBM along with enhancements to WebSphere and the System z and the growing embrace of SaaS by ever more vendors only reinforce Devine’s message. The Software-as-a-Service (SaaS) Showplace, an independent online directory of SaaS products, reports over 1300 organizations now have SaaS offerings.
Devine outlined the case for SaaS on the z in a presentation at SHARE last year, the SaaS-y Mainframe—Cloud Computing with System z. The core of the System z SaaS pitch, as stated at SHARE a year ago: Mainframes, as consolidation engines, are uniquely designed to virtualize and share everything—hardware, network, I/O, you name it—taking on the equivalent of hundreds or even thousands of servers workloads. Multi-tenancy, a core SaaS requirement, is baked into the DNA of the mainframe by default, which makes it particularly valuable for organizations re-architecting existing applications to deliver as a service.
Veteran mainframe data center managers were baffled when SaaS, ASPs (application service providers) and MSPs (managed service providers) first appeared on the scene years ago. That’s what they had been doing for years, for decades, they would tell me. Only, it wasn’t called that then. How is it any different from time sharing, they would ask.
Conceptually it isn’t very different. However, three things make it different enough: 1) the emergence of the Internet as a ubiquitous connecting fabric that everyone can use; 2) the browser as the universal client; and 3) the advent of services and service orientation. Previously monolithic code is now extracted as identifiable services and made accessible over the Internet via the browser following a requester-responder model.
Transzap, an online financial processor, remains IBM’s System z poster child for SaaS. The company handles payables, invoicing, and spend analysis through an Oracle database running as a service in a virtualized Linux on z environment. In 2009, the company handled 58 million customer transactions that way. You can check out their story here.
CA also has jumped into the mainframe SaaS game but from a completely different angle. CA wants to sell outsourced mainframe management to z shops as a contracted service. Their pitch is here. CA really is offering not SaaS but managed services more like MSPs offered a decade or more ago. Good luck.
IBM’s mainframe SaaS strategy envisions the mainframe as the center of SaaS offerings based on mainframe functionality delivered as sets of services. The goal is to enable IT to provide selected mainframe capabilities as online services and generate new revenue for the company. In effect, IT becomes what IBM refers to as rainmakers, using mainframe assets delivered as SaaS.
Are mainframe managers ready to think this way? Some certainly are. What mainframe data or functionality will your organization’s customers or new customers be willing to pay for?
In truth, the System z is well positioned to capitalize on the as-a-service phenomenon. The System z is multi-tenant to the max, which is critical to play the as-a-service game. Not only can it deliver SaaS data and functionality but also PaaS (Platform-as-a-Service) and IaaS (Infrastructure-as-a-Service). And it doesn’t take much to do this if you already have a z in place. Linux on z, WebSphere on z, and z/VM get you started.