Offloading z/OS Workloads
In the middle column, we start to see systems that could potentially be migrated off the platform but because they originated on z/OS, the System z team has a seat at the table to discuss the most appropriate platform. In this case, the resistance from the enterprise team was higher and the opportunities for Linux larger. That said, these workloads are still z/OS-oriented and are growing at a substantially lower rate than other areas of Linux.
Inheriting Mainframe Qualify of Service
In this column, we start to see workloads that may have originated on a distributed platform: These workloads may have nothing to do with z/OS at all. The rationale for using System z is that the customer wants to inherit some of the attributes of the mainframe. Most typically, this is disaster recovery, high availability or security, but might also include a need to consolidate servers to conserve data center space and reduce those associated costs.
The quantity, variety and complexity of these workloads are vastly higher than the previous three categories. Alas, so is the resistance to using System z as the platform of choice. Working out how to be more successful in this space will lead to significant success for Linux on System z. This is the target category that we usually think of when talking about “fit for purpose” decisions.
Implementing Enterprise and Cloud Linux Workloads
The final frontier in Linux deployment would be the general acceptance that the mainframe is a viable platform for any or all Linux applications. Organizations that have selected System z as the primary deployment platform and use it in this way definitely yield the greatest savings.
In a future article, I will highlight our conclusions for how we can remove barriers to increasing the deployments in categories four and five and how to create compelling events to encourage adoption. For now, it is essential that the System z community recognize there’s not only a big difference in the issues encountered, but also in the size of the potential market.