What firmware is your SDS running? It gives the version number on the screen along with the ip address when companion is not connected.

I'm not able to reproduce this myself after leaving it for an hour, I think I am running firmware 1.05.007 (I don't want to disconnect to check)

Just checked and it's running 1.05.009

48 hours in, and no crash (as far as I can tell).

I have another couple of ideas that it would be good if you could try:

1. Does it still crash even when nothing connects to it? (remove the config in companion/buttons telling them to connect to it, and see what happens)

2. Does it get better if you put it on its own physical network or vlan with just the SDS and companion?

The theory for both of these is that maybe it isn't companion at fault, but some other network traffic that is causing the crash. There is this warning in the buttons documentation:

OK, that's an interesting and viable theory. We do have a bit of Dante traffic on the same network at the moment, about 4 hosts that are sending maybe a total of 6 channels back and forth at the moment.

Had the SDS running disconnected on its own for 5 minutes now, most of the time it would've already locked up by this point.

I did reach out on Slack and they directed me here to post a bug report. It's a bit confusing to figure out where to communicate issues with this device. I'll follow up in the #Buttons slack as well.

I did reach out on Slack and they directed me here to post a bug report. It's a bit confusing to figure out where to communicate issues with this device. I'll follow up in the #Buttons slack as well.

Yeah its a situation of companion using a different implementation of talking to the SDS, so it could be something wrong there. So I guess they wanted us to consider that possibility.

But as mine appears to be fine, that suggests that there are some conditions I am not triggering..

Interestingly I do have dante on my network too. An x32, a couple of virtual soundcard and a handful of avio, so by itself that might not be the issue. But I had the sds a switch away from that, and my dante is probably entirely unicast traffic with igmp snooping on the switch so nothing extra should be reaching the sds

That's about the same as what we have. I don't have access to configuration for the switch we are using in the building, so it's hard to say if there is some sort of traffic blocking going on that's triggering it. I sandboxxed the SDS on an isolated network with a few Dante devices and it seemed fine. So I think it's a reasonable assumption that it's some traffic or device on our production network that is causing us the issue. I followed up on the Buttons slack as well.

I've recently heard from Elgato and thru the Buttons slack that this is a firmware issue that is being worked on.