I had a project to deploy Ruckus APs with a virtual controller on a network recently. The APs and their clients are on their own VLAN and get DHCP via Windows domain controllers on a separate VLAN.
I started setting things up, half of the APs got DHCP in the correct VLAN, but the other half got nothing at all. Wireshark shows no broadcast packets from the other APs arriving at the domain controller. The problematic APs DO get DHCP if we set up an isolated network with DHCP handed out by the switch.
We verified that the VLANs are all set up the same and working normally (a laptop plugged into the same ports gets DHCP as expected) and the scope is not out of addresses, and as far as I can tell there are no MAC conflicts.
What really boggles my mind is that the 'broken' APs consistently don't get DHCP, but the 'working' APs consistently did get DHCP from the same switches and switch ports. Even factory resetting a working unit would result in the same unit working again, but resetting the 'broken' ones wouldn't.
The domain controllers are in a Hyper-V failover cluster with several switches between them and the APs. All other clients on various VLANs and networks work perfectly. We're using all Brocade/Ruckus switches.
Ultimately we ran out of time to troubleshoot this today and I gave up and configured them with static IPs to get things rolling, but I'm curious if anyone has any ideas on where to look to figure out why this happened? All I can think is that there's some oddity like a misconfigured netmask somewhere but I couldn't find anything.
No comments:
Post a Comment