Nintex Automation K2, now with PostgreSQL connectivity.

Some assembly required. Results may vary.

I have been busy the last few months with a bit of self-indulgent experiment and I did a thing.

I have been encountering projects where PostgreSQL is often a part of the technology stack. It occurred to me that it was just a matter of time until I ran into a Nintex Automation K2 project that is going to need access to a PostgreSQL database.

So, I wrote a PostgreSQL broker, and I may have gone overboard.

Support row concurrency? Check.

Transaction support? Sure.

Want to still get partial support even if some column types aren’t supported? Why not!

This broker supports views, tables, materialized views and pretty much any out of the box PostgreSQL artifact that is discoverable and made sense to me at the time.

Supported artifacts include:

• ·Tables and partitioned parent tables

• · Views

• · Materialized views:

• · Functions

• · Procedures

• · Enums

• · Domains

Built-in utility artifacts: Diagnostics and Advanced Configuration ServiceObjects for compatibility reporting, discovery exclusions, generated SQL, performance history, and configuration validation.

Discovery is schema-allowlisted, permission-aware, deterministic, and non-executing. Unsupported or unsafe artifacts, such as anonymous record results, dynamically shaped routines, unsupported specialized types, or child partitions, are excluded or safely constrained with actionable diagnostics. PostgreSQL comments enrich artifact descriptions without affecting their stable K2 identities.

Schema-allowlisted is one thing I have done differently with this broker and is different from the typical K2 approach which exposes everything in a database, by default.

This is a bit of a paradigm shift, but in my experience, it has proven to be better to take a more conservative (granular) approach to service instance management and having to opt into schemas helps enforce better practice.

I am currently working on documentation, but I need some help testing it before I release it to the Nintex Gallery. I have also created an installer to make the process of registration of the service type easier. The installer is more of an additional experiment to determine if it is worth the extra effort upon which I would love to get feedback.

If anyone would like to help in starting the process of putting this broker through its paces, please feel free to drop me a note at postgres@caseitsolutions.com.