Systems and Operations, Companions to Code and Data Migration

by Stawsh (Stan) Murawski
February 8, 2010

When you move z/OS applications to a Mainframe Alternative (MFA) platform, there are more than only programs and data to move.  There are datacenter and application production procedures and policies and the operational practices in computer operations.  Depending on the MFA platform you have chosen, there might also be a significant new production environment to establish, with the attendant “systems programming” that maintenance of such a complex environment implies.

Establish the MFA System
Does establishment of an MFA platform imply MFA Systems Programming?  That answer is that it depends on the platform you chose.  To set up the production platform might be a minor or major, or even an incidental project.

To set up an MFA transaction processing environment, that fully and faithfully emulates CICS and z/OS Services underneath, can be as complex as the Systems Programming required to set up and configure CICS itself for the first time.  This task is made more difficult by the systems programming and production operations “learning curve” to acquire/learn skills unique to this new environment.  In this scenario CICS Systems Programming is not eliminated, rather it is migrated to, i.e., replaced by Systems Programming for the new environment.

If on the other hand the applications are transformed into the natural form of the target environment, production can be deployed into already supported production systems administered by existing personnel with mainstream skills.  For example if CICS applications are transformed into ASP.NET applications or into EJBs, then they run in a standard ASP.NET or Java EE environment.  In this scenario there is less to learn and less to install and configure.  Less to learn and do, and less to maintain, is easier, safer and less costly both in the project and ongoing

Run Daily Operations
I use the term application “run books” as a metaphor for all that IT Operations is responsible for about running the applications and responding to any failure to run.  The important and enduring legacy of the mainframe is that of the excellent service that has been provided to the enterprise or organization by mainframe IT shops.  However the real source of the quality of this service, e.g., its reliability and security, is the way that a good mainframe shop consistently follows good processes using proven procedures.

While we might hope that the “rules” by which the shop is run would be documented, often they are not.  While well codified and consistently followed, in a long established shop much of the process is what might be referred to as “tribal knowledge” embodied only within the current staff’s heads.  If the operational workflow flow and event conditions are “documented” anywhere it might be in the current “job scheduler” settings and specification that came out of the historical “run books”.