For years, the mainframe has been a reliable workhorse in corporate IT environments. New hardware and operating system technologies, combined with comprehensive management strategies, have ensured its continued ability to satisfy the business requirements for high-availability and cost-efficient application hosting.
The statistics speak for themselves:
- More than 70 percent of corporate bet-your-business data continues to be hosted on mainframe systems, according to a study conducted by Hurwitz & Associates (see Hurwitz & Associates, “The Mainframe Software Market,” Robin Bloor, 2006).
- More than 60 percent of mainframe MIPS installed since 2000 are being used to process new application workloads, according to research conducted by Software Strategies (see “Top 15 Reasons Users Should Stay On/ Upgrade/Move on to the IBM Mainframe,” Ian Bramley, December 2006).
- More than 800 business applications are now available on Linux on System z mainframes.
It’s clear Big Iron remains at the forefront of enterprise computing. Despite two decades of predictions to the contrary, the mainframe platform is more important than ever.
The mainframe has come a long way since the ’80s, when it was cloistered away in some data center enclave. It’s now a fully connected resource in the distributed, Web-enabled enterprise, where it carries ever-growing, mission-critical business functions. The IBM System z9 mainframe’s new hardware and operating system technologies, complemented by innovative, Independent Software Vendor (ISV) management solutions, further boost its capabilities and performance to satisfy the dynamic and challenging requirements of today’s IT-centric businesses.
Here are 10 reasons why the mainframe is growing in importance:
1. Security: External attacks and malware threats are virtually unknown to mainframe users. This unparalleled intrusion resistance is achieved through robust, time-tested security solutions that have been designed to exploit unique properties of the mainframe hardware architecture. These solutions have been an integral part of the mainframe ecosystem since its inception but have only recently reached some level of maturity in the distributed world.
2. Investment protection: Applications developed decades ago can run unmodified even on the newest mainframe models. Moreover, they can readily coexist with newer Linux- and Java-based applications. It’s even possible to integrate these core business applications into today’s Service- Oriented Architectures (SOAs) —presenting their embedded business logic as addressable business services.
3. High availability: The mainframe’s design, self-monitoring capability, and advanced high-performance clustering (i.e., sysplex) let it satisfy the highest “five 9’s” availability requirements. Applications can run reliably on the mainframe for years without any unscheduled disruptions. “Bulletproof ” is a key distinguishing property of the mainframe that makes it the de facto standard in reliability against which all alternative platforms are measured—hence the use of popular phrases such as “near-mainframe availability.”
4. Workload and performance: Mainframes run 90 percent of the world’s largest databases. There’s hardly a more compelling argument in support of its performance characteristics. The mainframe also is highly efficient. z/OS facilities such as the Intelligent Resource Director and the Workload Manager ensure near 100 percent utilization of all mainframe resources. Based on a 24-hour observation window, comparable values for Intel/Windows or Unix/ RISC servers are 5 to 10 percent and 15 to 20 percent, respectively. Newly introduced specialty processors such as the System z9 Integrated Information Processor (zIIP), Application Assist Processor (zAAP), and Integrated Facility for Linux (IFL) improve the overall cost efficiency of the platform by offloading selected work from the much more expensive general-purpose engines, thus creating new “white space” (unused capacity) for customer business applications.
5. Efficient interoperability: The mainframe has the unique ability to efficiently co-host multiple disparate systems and workloads while maintaining high availability and quality of service. This makes it an ideal platform for server consolidation. Multiple Linux and IBM z/OS operating systems, for example, can be hosted on a single mainframe box. Mainframe-only technologies such as HiperSockets can provide direct and secure memory-to-memory communication links between any of the Logical Partitions (LPARs) hosting these systems with near-zero latency. A Java-based application running under Linux in one LPAR, for example, can access an IBM DB2 database hosted by z/OS running in a second LPAR with virtually no latency in the communication path. That’s something no other platform can do.