We as an Integrator are new to using Ignition and would like to know if anyone has experience with our redundancy scenario.
We have 2 servers set up as a redundant pair each with a PostgrSQL database running. If the connection to the database failes on the master gateway - we would like the backup to take over and become active.
I suggest you handle database redundancy external to Ignition. Host the database on its own server(s), using your database of choice's replication/redundancy mechanism. It should appear and be configured as a single connection to Ignition.
This is how a customer of ours does it with SQL Servers and a Availability Group. They create one "listener port" that Ignition connects to, so we only have one db connection to it, but behind that connection port, they have a redundant pair of database servers that failover as needed.
PostgreSQL can do replication natively, but it doesn't have built-in tools to manage failover. Theoretically, you could script it all yourself with HAProxy or something similar, but repmgr and barman are well-known and supportable tools.
(EnterpriseDB offers support for PostgreSQL and these tools. EDB is the largest employer of contributors to these projects, IIRC.)