Sometimes various events may cause the bridge link to fail. The following tips may assist in troubleshooting the link.
While connected to the LAN port of one of the PoE injectors see that both ends of the link can be PINGed. Start with the nearest unit and then try to PING the far unit. If the ping to the far end fails then repeat the PING process at the opposite end. This can indicate if either AP has been reset, is failing or has lost link. If the ping to the far end succeeds then the link is up and there may be an issue with mild interference or not related to the link itself.
The most common reason for not being able to ping the units when they're in their preconfigured state is because the workstations on the client's site do not have an IP address in the same subnet as the access points. The APs in our preconfigured kits are intentionally on an uncommon subnet with IPs of 172.17.2.100 and .101. A workstation will need to have an IP address in the same subnet to be able to ping and access them.
Check the alignment of the APs to ensure they're pointing directly at each other and that there is clear line of sight and fresnel zone between them.
Changing the frequency of the master AP to a quieter channel or adding RF shielding can help to avoid environmental noise.