I inherited an industrial setup with 20–30 controllers mounted on moving booms. The controllers connect by Ethernet to TP-Link TL-WR902AC units configured in client mode, and those units provide the Wi-Fi connection. Unfortunately, every TP-Link still has its default hostname, so it is difficult to tell which device belongs to which boom when one has a problem. The MAC labels are covered by industrial adhesive, and the units are mounted inside secured boxes about 10 feet up. They also do not appear to respond to CDP or LLDP while operating in client mode. I know the controllers and access points have IP addresses, but I would prefer to identify the bridges without repeatedly climbing up or disrupting the equipment. Is there a practical way to map each bridge to its MAC address or physical location?
4 Answers
If a brief interruption is acceptable during a maintenance window, reboot one bridge at a time and watch which controller or wireless client drops offline. That gives you a reliable physical mapping without needing to read the covered labels. For a completely non-disruptive option, you could also walk around with a temporary access point and use signal strength to narrow down the location of each wireless client, although that is more work.
You can identify them without rebooting anything by generating traffic from one controller at a time. For example, start a ping from the controller on a particular boom and watch the access point association table to see which wireless client becomes active. If the controllers are reachable by IP, check the ARP table as well; it can map controller IP addresses to MAC addresses, and the controller's addressing should tell you which boom it belongs to.
LLDP generally will not cross a Wi-Fi client connection, since it is an Ethernet link-local protocol and wireless bridging behavior varies by device. If you can log into the TP-Link units, check whether LLDP can be enabled on the Ethernet side, but I would not depend on it in client mode. In practice, the IP/ARP mapping or a controlled reboot is likely to be more dependable.
The ARP and DHCP tables on your firewall or router are probably the easiest starting point. Make a table containing each controller's IP and MAC address, then correlate that with the TP-Link client entries. Once you have the mapping, change the hostnames so this is easier next time. Keep in mind that the Ethernet MAC and Wi-Fi MAC belong to different interfaces, so plugging a switch into the LAN port may show a different address from the one listed in the wireless association table.
I do have the controller and access-point IP addresses, so I can continuously ping the controllers and observe which one disappears when a particular bridge is rebooted. I can do that while the booms are parked and then record the IP and MAC mappings.