In distributed development, application programming interfaces (API) have been easing software development innovation and product integration for a while. The mainframe must start providing similar APIs to fit with the larger software ecosystem.
For those in the mainframe space who are unfamiliar with REST APIs, these are web services that connect mainframe programs and services. REST APIs are based upon HTTP protocol and provide a simple way to remote-connect a client using a REST API to a service provider.
REST (in the running for one of the worst acronyms of all time) stands for representational state transfer. Basically, it allows you to move “state” information from your web service to your client.
In recent years, REST APIs have become standard building blocks that make it easier to develop programs. Developers use REST APIs to embed Google Maps on webpages, integrate YouTube videos and functionality into websites or applications, and gain programmatic access to Amazon’s product selection and discovery functionality to advertise Amazon products and monetize websites. You can see a growing list of over 15,000 APIs here.
The Need for REST APIs
For years, distributed development has continuously lapped mainframe development with rapid innovation. But the mainframe’s historical relegation to a slower operating mode hasn’t obviated its vital role in digital transformation, and it’s a lie that the platform is too out of shape to pick up speed.
As the most reliable and securable platform available for handling hefty loads of web, mobile and data transactions, companies must continue fostering mainframe agility and expanding the mainframe-inclusive DevOps toolchain to help it get on par with other platforms.
Mainframe REST APIs provide the most direct way to integrate mainframe services and data into your digital strategy. As mainframe REST APIs become more prevalent, they will allow developers without specific mainframe skills to easily access data on the platform.
For instance, mainframe products that provide REST APIs allow visibility into what’s occurring on the mainframe without having to actually be on the mainframe. Web-based API’s allow remote access from any other platform and make the mainframe look like just another, albeit bigger, server on your network.
Enabling Integration, Automation, Innovation
In mainframe development, if someone is unavailable to perform a manual task, it could be days before a process is completed. A big advantage mainframe REST APIs provide developers is the ability to tightly integrate products into their processes and eliminate manual steps and hand-offs between groups, which cause time lags.
Replacing manual steps with mainframe REST APIs accelerates the development processes. It’s as simple as calling the API from a piece of automation software and using a scheduler to start it, or using a notification from your process to initiate it in your automation software.
At a high level, mainframe REST APIs allow developers to:
- Connect multiple products: Using APIs to connect multiple products allows you to merge data from different sources to provide a better context for solving a problem. Combining data for more products can provide identification of a problem and point to a solution. For example, you may have a marketing campaign where you need to find demographic information from the census data near your customer’s zip code. You can access the census data and your customer data with a REST API.
- Capture user expertise: With mainframe REST APIs, you can capture expert usage knowledge from a product and institutionalize it in an automated process. Your experts can capture the necessary steps to get data from a product into an automation tool or script. Once these steps are automated, your expert’s can use their time to learn new skills that can be rolled into the next wave of automated processes. That way experts can focus on advancement rather than repeating the same manual tasks. For example, your best performance analyst might pull data about a specific set of transactions using a REST call and also pull data from a specific report that indicates whether or not the transaction program needs more memory.
- Scale workloads: Because APIs enable the automation of product usage, they also let developers scale the level of work beyond what people executing manual steps in a product can accomplish. Scaling across more systems through automation allows your team to handle significantly more systems than they ever could manually.
Agile culture, processes and tools were once exclusive to distributed development, but today Compuware has fully implemented mainframe agility and proven the mainframe can function in the same Agile DevOps environment other platforms thrive in.
Similarly, while APIs have become standardized in distributed development, they’re also becoming a vital part of the mission to mainstream the mainframe. In following posts, I’ll begin discussing how mainframe REST APIs apply to specific steps in the DevOps process.