Pronouncements of the death of the IBM Mainframe started in the 1990’s and continue to this day, with CEO’s vowing “to get off the mainframe”. They lack the understanding that the mainframe is the most important server in their infrastructure. At the same time, z/OS applications have become the systems of record for customer facing applications. Customers transfer money, check bank balances or find out where their package is through applications that use System Z on the backend. These applications have caused an exponential increase in the number of transactions being serviced.
MIPS Management techniques from yesterday are no longer sufficient to address performance and remediate costs. Today users need to understand the Rolling 4 hour Average (R4HA) MSU usage. When is the system is at its peaks, and what is running during the peak. Specialty engines help mitigate these MSUs, but what is able to run, and what is able to run but didn’t run, is important to understand.
IBM recently announced a new program around the Mobile Workload that allows companies to deduct 60% of their Mobile MSUs from the R4HA potentially leading to reduced monthly bills. Identifying the mobile workload is going to be problematic initially. Simply gathering all your SMF 110 records and saying that is your mobile workload is probably not going to work. For example, a customer using a desktop browser checking his bank balance would not qualify, but a customer using a mobile browser on their smartphone to execute the same activity does qualify. Organizations are going to need to draw a distinction between the two as well as 3270 access. The Mobile “Offload” could save significant money for an organization -- or not -- depending on when the R4HA peak is.
The speaker will discuss these topics and how new strategies around MIPS Management in a Mobile world can save costs as well as improve performance.