Mainframe workload rehosting is increasingly popular. But the traditional approach of requiring the recompilation of all source code is fraught with issues — issues that are not apparent at the whiteboard stage of any project.
The arcane legacy of 50 years of mainframe programming, has created a bewildering array of barriers to success for any project that relies on source code – barriers made insurmountable by the skills shortage at large within many users of legacy hardware mainframes.
This white paper examines some of the less obvious issues and the impact they can have on a workload rehosting project based on source-code recompilation, such as:
- Source code availability
Source code management in the early pioneering days of commercial IT relied heavily on local system knowledge. The source code management disciplines we take for granted today only became commonplace as the job market for programmers exploded and staff turnover increased. This means that organizations find it difficult to identify which source code was used to build aspects of long-running applications.
- Application interdependencies
Once a user starts looking, it's very common for a list of program dependencies to grow exponentially. A program may reference two other programs, which may in turn reference another two, which may be called by several other programs. Each of those programs may have their own dependencies. In even a modest system, this Gordian Knot of interconnections extends the scope and duration of a workload rehosting project.
- Mainframe encoding vs x86 encoding
Mainframes and x86 computers encode characters and data in different ways. While this isn’t a problem with storing the data, this can cause serious complications with program logic. For example, when applications deal with money, mainframes maintain a unique representation of decimal data called "Packed Decimal", which has no equivalent in modern data encoding. Integers, too, are represented.
- Data processing and databases
Modern computers perform floating-point operations differently than mainframes; the maths yield a different result. For certain classes of applications, failure to account for these differences can have a very serious impact.
- Obscure mainframe coding techniques
For decades programmers have been utilizing system-level features of the mainframe environment in programs to circumvent operating system limitations. These arcane routines become impossible to maintain once the original programmers leave the company, and creates an almost insurmountable lock-in to the mainframe operating environment.
Location and documentation of the source code is one issue but, as you see, there are other challenges that need to be overcome to manage a successful mainframe workload rehosting project.
Learn more about these issues and how these challenges can be mitigated. Complete the form to access the white paper.