CAP service enables mainframe

Keywords:

Mainframe, REST, 3270, JSON, Robots

The context:

A company has a requirement to use data accessed via a mainframe terminal interface in a process

automation solution.

The solution:

Using CAP, the mainframe access was emulated using the 3270 Extension and made available to

other systems by exposing a REST API using JSON directly from with a CAP Agent instance.

Deployment diagram:

Why CAP:

CAP is capable of interacting with a vast array of data sources, protocols and APIs out of the box. In

this case, the built in 3270 terminal emulation for mainframes along with the built-in capability to

expose a service, made CAP the obvious choice.

The story:

Upon receipt by the business of an account closing request, many systems (including the mainframe)

need to be checked to verify that all outstanding debts and balances are cleared. As a manual

process, this takes up to an hour per account. So replacing this check with an automated one that

only flags exceptions is highly desirable. However, the IT option for creating a service directly from

the mainframe was deemed too costly to justify in the context of the benefits from the automation

project.

Using CAP, a simple screen-scraping option was used instead. Upon receiving a HTTP POST with a

JSON formatted data request, CAP will check if an existing session to the mainframe is present. If

yes, it will simply perform the required transaction codes to check for balances.

Should a mainframe session not be present, CAP will log in first, before responding.

To ensure that the mainframe is unable to be flooded with requests and to only ever keep one

session running, queuing rules were added so that at any point, requests would simply queue up and

wait for the session to become free. This approach was deemed satisfactory as the average number

of transactions per day was below 100.

Once the service was made available, a second CAP instance was used to create the end-to-end

process of checking all other systems as well as the mainframe. This ensured not only the success of

the automation project, but also made the service available to other consumers within the business

should the need arise.

Rules blocks and components used:

3370 Emulation Rules

Queuing Rules

String manipulation Rules

JSON Rules