CLU 1.1.2 shows nodes at first, but then refuses to find them unless powercycled

NETRON
1

New to this hardware, not to entertainment networking. CLU 1.1.2 on Windows 11. For setup purposes I am connected directly to an NS8 on 1.0.0 which is feeding (2) EN4s on 3.0.1. All Windows firewalls are off. When first powered up, the CLU sees all three devices. After a number of minutes (10?) a refresh shows NO DEVICES AVAILABLE. However I can still access them through Chrome by navigating directly to their IP addresses. Power cycling the nodes allows the CLU to find them again, but the behavior is repeatable and consistent. IP and subnets are compatible in the 10.101.x.x/16 range (though I’ve seen the CLU is able to see nodes even if they’re incompatible). I believe unrelated, but the only tweak I’ve made beyond basic IP/name stuff is changing the IGMP querier from auto to custom and assigning the querier IP to the switch’s own IP. However this behavior occurs whether the IGMP settings are on auto OR customized.

UPDATE: I realized that the node was dropping sACN transmission around the same time, which made me revisit the IGMP settings. I’d failed to put in values for the query intervals, maximum response time, etc, so I added them. This brought the sACN back on line, but now after 10 or so minutes the CLU fails to see the NS8 (though it DOES see the EN4’s downstream now) BUT I CAN NO LONGER EVEN ACCESS THE NS8 THROGH ITS IP INTERFACE. I can also no longer even ping the NS8 at its address?! Running wireshark shows the NS8 is still alive at its IP (10.101.170.111) but has no traffic at all beyond a group query every ten seconds. It passes traffic but there is no longer a way to interact with the switch at all. Again, a power cycle brings it back to the CLU and the IP interface, but that’s a hideously not ok solution for something going into a lighting rig.

The forum won’t let me embed two images, so I can’t share the wireshark demonstrating the NS8 is, in fact, online at IP 10.101.170.111, but only sending a group membership query every 10 seconds and nothing more.

Below is ping results (‘.111 is the NS8, ‘.112 is one of the EN4’s downstream of the NS8):

Can ping downstream EN4 but not NS8 the data passes through
Can ping downstream EN4 but not NS8 the data passes through1182×697 18.3 KB

Someone please help me out here. Surely I’m missing something?

Thank you.

2

Further update:

My ProPlex IQ tester is unable to see any sACN traffic through the NS8 when the snooping is custom set (querier = the switch’s IP, all other timings and groups = 10) EVEN THOUGH the NS8 is accurately passing sACN to its downstream EN4s. I have used the ProPlex on countless rigs for the last 3 years across many manufacturers and never seen it fail to accurately see sACN and the sources. However, if I turn off snooping entirely (the only place I see to actually toggle off IGMP snooping is in the VLAN menu, where I can toggle off both snooping and queries) then the ProPlex is able to see the signal. I would accuse the ProPlex of not correclty subscribing, but it has worked flawlessly on other manufacturer’s equipment (ie Luminex and Pathway). I’m not sure what to make of that fact.

I’m at my wit’s end with these things. Please someone show me what I’ve misunderstood. Thank you.

3

Hi @Luke

Thank you very much for your detailed description and troubleshooting steps.

CLU uses the LLRP protocol for device discovery, which utilises multicast addresses. Therefore, incorrect IGMP settings can affect LLRP. Are you still seeing these issues, with the NS8 in its default IGMP settings…?

  • IGMP Querier enabled per VLAN
  • IGMP Snooping enabled per VLAN
  • IGMP Configuration = manual, with values reset

The default IGMP Query Interval for the NS8 is 20 seconds. This has been chosen after thorough testing with multicast protocols used on entertainment networks. This is a fairly aggressive IGMP query interval - many IGMP queriers feature a query interval of nearer 2 minutes.

The IGMP Query Max Response Time must be set less than the IGMP Query Interval. A Query Interval of 10 seconds is very aggressive. Therefore, unless there is a very specific need, I would recommend setting the Query Interval to at least 20, with the Query Max Response Time around half the Query Interval. These are the values that will be set upon clicking the Reset button in the IGMP Configuration webpage.

Please let us know if you have any questions.

Edward