Cost is the primary driver for shifting COBOL compute to less expensive platforms. But once organizations decide that they are going to use shifting COBOL compute they discover there are other benefits to this modernization method.
These benefits are realized in three areas: impact, risk, and momentum.
The immediate benefit of shifting COBOL compute can be seen in the reduction of mainframe MSU utilization of the environment from which the program or application has been shifted.
The reduction will be seen by comparing the Sub Capacity Reporting Tool (SCRT) listings before and after the shift has been completed. The SCRT captures MSU utilization and calculates the rolling four-hour average (R4HA) of the environment (e.g., LPAR). The R4HA is used to calculate the Monthly License Charge (MLC) for IBM software, which is the majority of the cost in the mainframe environment.
By lowering the R4HA, the organization can achieve lower MLC. Lower MLC should result in a lower cost of executing the application or program.
Additionally, since shifting compute utilized compiled Java bytecode, the legacy application can be maintained using traditional methods. The programs remain in COBOL and are cross-compiled into Java, not maintained in Java. Existing support staff are not required to maintain the application in a new environment resulting in minimal impact to the current team and staffing levels.
Applications that fit the profile for shifting compute (i.e., must reside in COBOL, are computational heavy) are often mission-critical and involve the organization’s intellectual property. They include proprietary calculations, processes, and data movement activity. These applications provide the business with a competitive advantage in the market. Disruptions or failures for these applications are not acceptable.
An emphasis on legacy data compatibility can be utilized to ensure data equivalency (formats, calculations, and results) is achieved. Using existing Db2, VSAM, and QSAM files minimize changes to upstream and downstream integrations, eliminating the need to change other applications or programs, and helps achieve functional equivalency.
Shifting COBOL compute is a low-risk modernization method, but that does not mean risk does not exist or cannot be mitigated. Data and functional changes (non-equivalency) is the risk that can be mitigated with automated transformation and testing.
Automated transformation ensures consistent and predictable Java outputs. Configuration of the automation engines helps to improve the execution performance of the new Java executable, boosting the drive to reduce disruption risk.
A successful shift compute project can help an organization establish modernization momentum.
Success breeds success, and when cost reduction and low-risk modernization efforts achieve their desired outcomes, it will encourage additional projects. Additional projects could achieve additional cost savings. Those savings (budget or real dollars) can fund other modernization efforts.
It’s also important to note that not everything has to change at once, meaning an incremental approach to modernization can be used. The organization may begin with one program within one batch job stream and then expand.
There are clear benefits to shifting COBOL compute. Some of these benefits are tangible (reduced costs), and some of them are intangible (momentum and incremental modernization paths).