Vision Panel and full Ignition licenses

We are considering putting a Vision Panel license on each client because we were told that if the client server dies, the client can still operate as a stand-alone, less a DB. My question is this;

In order to operate as a stand-alone the client would need to have Ignition installed on it so it could run a project on its own. Development and maintenance are performed from the main Ignition server. So, if all modifications are performed on the main server, how do the modifications get from there to the project on the client that would be used in the event of a server failure? Do we have to always u/d the project and, exp/imp the tag DB to the clients Ignition? :scratch:

The key point is that this architecture style typically uses the retargeting feature to run, manage, and maintain the project centrally - leveraging, in my opinion, the strongest feature of the Ignition Platform. While the network is up, everyone runs the latest version of the central project and gets the benefit of the fully licensed server. Those rare network or server downtime events are the only time that you utilize the panel license locally - to exercise control. With this model, you shouldn’t need to make frequent modifications to each node. If you do - they are stand alone by design - think legacy RSView. That means that each panel station necessarily needs to be updated separately (although you can run the designer across the network and might be able to template projects). I’m not sure why you would need to frequently modify tag databases.

Panel nodes don’t factor into dealing with project changes upon a server failure - they exist for local control. You can mitigate server failure by using clustering, which load balances clients, stores project changes, and gives you a higher degree of server reliability. Besides that, you can do frequent backups, or even have a “cold” standby server. Having a reliable network and database helps greatly here as well. These features all complement the Vision Panel model - just keep in mind that panel nodes and server reliability are separate issues.

While you could run a true “peer-to-peer” setup - you usually wouldn’t want to.

Nathan,

Thank you for the reply. I just wanted to get it straight as to the capabilities, anomolies, etc… (didn’t want to over-promise, under-deliver to a customer). It is still way better than the others out there! We thank everybody at IA for their great work.

We do have plans for a way of automatically syncing projects to stand-alone vision panel edition nodes, but it won’t be available for a while. So yes, in the meantime, you’d manually update the panel editions with their control-only version of the project.

In conclusion: Could you set up a cluster (of 2 servers) with one works and one panel ed. server?

Yes, you could do that. Often you’d have more than one panel edition for a given “Works” gateway, but if you only had one then you could do that. The clustering in 7.2.0 would work well for this with its more well-defined “master” and “backup” definition.