Transform Engine Continuously Disconnecting and Reconnecting

I have a client with a transform engine that is having connection issues. Yesterday it was disconnecting and having to reconnect every 30 seconds to a minute. We have two separate connections to the engine, from both the control port and the Dante primary port, both are on the same subnet as the Mac Mini being used for control is being used for DVS as well.

We are using Netgear AV-Line switches, M4250 in a star configuration over Fiber Lags to an M4350 series core switch and using the Dante Audio profile, switches are on recent firmware as of two weeks ago.

Any help or direction would be appreciated.

The first thing I’d change is having both the Control port and the Dante Primary port in the same subnet. The engine is then reachable for control on two IPs in the same broadcast domain, and the client can end up bouncing between them.

Quick test: unplug the Control port and run everything over Primary. A single Primary connection carries control as well as audio. If the disconnects stop, that was it.

If you want to keep the Control port in use, put it on a separate subnet/VLAN and give the Mac a second interface in that range.

Thanks for the reply. They were having this issue before but had been running with only the Dante primary port in use for both audio transport and control. I talked to G1 and they recommended using the control port for control, but the Mac mini that’s being used for control is also being used for DVS hence them being on the same subnet.

I can try and move the control port to the control subnet we have and connecting via the Mac’s WiFi nic to keep it separated. But connecting via the control port was supposed to be the fix.

Hi all,

As this was happening with only Dante Primary connected, I’d suggest simplifying the network and working back from there.

Start with a point-to-point connection between the Mac and transform.engine, then progressively add the switches/netork devices back into the path until the instability reappears. That should help identify where the issue is being introduced.

I’d still recommend keeping Dante Primary and Control on separate subnets/VLANs if both are connected. Also, the transform.engines control port is 2.5G so this could alleviate any possible bandwidth issues.

Let me know how you get on.

Ross
Product Specailist

Since it also happened with only Primary connected, I’d treat the dual-IP/subnet issue as a likely aggravator rather than the root cause. Still, I wouldn’t use the Mac’s Wi-Fi for the control port; Wi-Fi roaming and power saving can create their own intermittent control drops. Use a wired second interface or a VLAN on the existing NIC instead. Then run Ross’s point-to-point test and add one switch at a time, checking Dante Controller for clock leader, latency and error counters, plus IGMP snooping/querier on the M4250/M4350.