ePMP 1000 GPS; ABAB frequency reuse(2412 and 2467 20MHz); 75/25; 4.4.3
One AP went crazy last night for no apparent reason. Knocked nearly -ALL- SM’s off all 4 APs (I am guessing sync and transmitter spurring issues due to knocking SM’s off 2 AP’s not on the same frequency as this AP). Problematic AP also dropped its ethernet sync to 10Mbps for no apparent reason.
Upon reboot of problematic AP, all 2.4Ghz ePMP SM’s on tower began to reconnect to their respective AP’s. But problematic AP was not accessible. Netonix showed data going to AP but 0 data being received from AP. AP could not be pinged.
Upon arrival to the tower, tested AP straight into a PC via POE injector. AP tested fine.
Also noticed 4.4.3 was now in inactive bank and 4.4.2 was active. Possible firmware corruption?
Reapplied 4.4.3. and now both banks display 4.4.3.
Tested the original port on the Netonix this AP was plugged into. Data out, 0 data in. Tested with 3 devices, same issue with all 3, port is definitely bad. The weird things is the port will sync up at 1Gbps, just no data. Syncs up at 1Gbps with the AP at the top of the tower as well, just no receive data.
Moved AP to another port on the Netonix. Powered it up and been working every since.
These AP’s are at the top of a 240’ Rohn 55 with 2 FM translators (700w and 250w). We have ferrite cores on our other AP’s and backhauls but not on the 1000 GPS AP’s (never had any trouble with FM translators bothering ethernet sync for the 1000’s). There is also a lightning arrestor above the top of the tower (10’ ground rod).
Tower and radio shack has a ground ring with single point ground inside the shack. Ground system is in very good shape. Cambium 30v suppressors are grounded to a large copper buss bar inside shack (all equipment inside the shack ground to this single point). No suppressors at top of tower, but radios are grounded well to the tower. We did have a bad storm season this year, so not ruling out a ‘touch’ of lightning earlier this year that just manifested itself.
The only recent changes made to this tower was a PTP 550 link was upgraded to 4.6-RC35 in hopes to stop some intermittent link drops using channel bonding (4.6-RC35 did fix that issue). That was very early Saturday, over 2 days ago.
Has anyone else seen anything similar to this and hopefully know exactly what caused it?