I would like to get the community's help to know the redundancy architecture's practical implementation,
We have six critical machines that need to connect to the ignition server.
Since the machines are critical, we are planning to propose the redundancy architecture at the server system level, which will have ignition and a database installed in it.
Queries :
Will Ignition work as a Hot redundancy? Will It support bump-less transfer from primary to secondary servers for the application, device communication, and database connections etc.?
How much time is required for the switchover?
If bump-less transfer is not supported, what are the limitations of the redundancy? What are the manual interventions and actions required during the redundancy action?
How will ignition work in the case of Windows failover clustering being implemented for databases? (Not shown in architecture).
Recently implemented redudnancy for a 8.1.26 server on hot redundancy.
My application is only a database with minimal devices so really only the db connections gets swapped. The vision clients recover in under 10 seconds (usually faster). I think technically its a "warm" not "hot" for Igniiton, so it may not be exactly what you are asking. Worth reading through - Setting Up Redundancy - Ignition User Manual 8.0 - Ignition Documentation
How will ignition work in the case of Windows failover clustering being implemented for databases? (Not shown in architecture ).
Meaning your Ignition connects to an external SQL Server (not on the same computer hosting Ignition)? I actually have an ongoing issue with this. My setup is that our Ignition servers connect to a SQL Server listening port as part of a SQL Availability Group Cluster (and behind that port it's 2 different machines hosting SQL Server). Sometimes, its completely fine, but a few times, we've had to restart the gateway to re-establish the db connection. I actually have a ticket setup and observing the failover today so I might have more information for you regarding this later today. Will only be helpful though if you're setup is similar to the one I described ie both Ignition servers connecting to a listener port as part of a SQL Availability Group Cluster.
