I've read through the forum and seen a number of sophisticated setups to handle VMix Call in case of failure.

I'm looking at a more basic setup, and was wondering about some details.

We will have two VMix systems with different operators in different locations / networks. The show has a total of 6 call-ins.

Our plan is to run the show of VMix system A, with VMix system B following the show flow as a warm standby (meaning ready to go, but stream is not active).

If system A fails, a brief TD slide is shown on platform and we bring system B online to take over.

I'm not clear if the same VMix call IDs/URLs would work in that case, or if there is a risk for stale information causing routing failures. Our plan was to reconfigure system B with it's own set of call IDs/URLs by switching the input details after the project is cloned for the second system. Then provide both sets to the talent. If they have to reconnect due to failure they would use the backup set of IDs/URLs.

Not as elegant of course as anything that's more automated. But we're keen on keeping it simple with a tolerable experience.

Thoughts appreciated.

Jan

That makes sense. But in that scenario, VMix B can't be open (running) until the point of failover, right? The issue would be that with call manager open on both systems with the same IDs, how would the call protocol know whether to connect the call to VMix A or VMix B? Would it chose randomly? I'm operating on the assumption that the VMix Call protocol doesn't support being connected to two VMix call servers at once.

You could leave VMix B closed until failover is needed, but then you have to start it and operator for VMix B needs to get settled. That lengthens the failover time.

In our scenario VMix A and VMix B were in different cities with separate operators, not two systems in the same place.


Since posting the question, we did run our show. I built VMix A. When it was ready I saved it as a preset. Opened the preset and just changed all the Call IDs for a preset for VMix B (simply go into 'change' on the input, which automatically assigns it a new ID). Then we copied all the files and the preset for VMix B on the second system. Worked fine. And the network gods were with us, so we didn't have to failover, but were prepared.

Even though the talent got well marked sets of links, one caller first clicked on the backup link, and we had to tell him to use the primary link. But that was an easy fix.

Thanks,
Jan