IBM Systems Magazine, Mainframe - March/April 2017 - 31
supports that need. But others are interested in delving into digital innovations. They're using DevOps and continuous delivery models, which IBM supports, as Linux is the reference platform for these innovations, Adlung notes. DevOps embraces and uses container-based technologies and builds on microservice architectures. All of these are available on Linux on z Systems. "Clients can use those technologies one by one just as they would on distributed platforms," Adlung says. IBM's three distribution partners provide DevOps tools that are popular with developers and available on z Systems. In addition, IBM is constantly receiving feedback from clients that bring new workloads to the platform, Thoss says. Digital innovation encompasses cloud, too, with the development of next-generation, cloud-native apps. DevOps capabilities using Linux on z Systems help make developers' lives easier and give your company a competitive advantage, Hosch says. Workload Efficiency Every organization relies on its IT platform to handle diverse workloads; this is another area where mainframe excels. Linux on IBM z gives the data center the simplicity and efficiency of a single server that's scalable and responsive. The reliable platform can run mission-critical operations as well as varied workloads such as analytics and blockchain. Multiapplication workloads can be handled effectively when Linux is employed on z Systems. Analytics workloads have become central to every business. With z Systems, data isn't pulled from another server and replicated. Instead, it resides close to the workload, thereby reducing latency. Data lakes that exist in a distributed environment can't exist in a z Systems environment. "We show clients how they can host their data and analytics in close proximity to change their SoRs into systems of insight," Adlung says. Any client that wants to run analytics as a service can do so using the mainframe. For a number of years, IBM has supported Apache Hadoop and Spark open-source analytics software offerings, which run on Linux. IBM also has enabled newer open-source analytics solutions such as MongoDB and PostgreSQL, Thoss notes. Analytics influence not just software but also the OS and hardware. "It has to be part of the stack," he says. Analytics rely on data serving, and z Systems hardware boasts fast processors, large caches and dedicated I/O coprocessors to handle these data-intense workloads. The platform's design gives users the ability to implement more complex analytic algorithms to gain better insights, faster. A client that recently evaluated moving its fraud detection neural network onto Linux on IBM z described the latency reduction as "opening the aperture on innovation," Mitran says. "Such improved insights help our clients make better business decisions while enriching the end user's experience," he remarks. Mainframes enable tailored solutions-such as blockchain- to be created. Blockchain technology has been available for many years, but the commercial interest in it is recent, Adlung says. For clients looking to build their own blockchain environment, IBM can assist with blockchain on-premises in the data center. Through Bluemix* technology, IBM also offers access to blockchain services in the cloud so clients don't need to host them in the data center. In January, the company began offering an on-premises ibmsystemsmag.com MARCH/APRIL 2017 // 31 pg 28-32.indd 5 2/7/17 1:43 PM