APs unstable above 2.4.2

I upgraded all of my epmp, about 700 radios now to 2.4.3 and started having issues with client radios intermittently not responding. I would have blocks of radios stop reponding at the exact same time. If we would log into the AP the non responsive radios would show that they were in session. If I de-register them once they reconnected they would work again. The issue would show itself randomly on various towers and spread amongst multiple access points on the tower. Once I downgraded back to 2.4.2 on the APs the problem went away. As soon as I upgraded to 2.5 I immediately started seeing this issue again. I have already started downgrading my access points back to 2.4.2. but there is obviously something different above 2.4.2 on the access points that makes it unstable and therefore unusable for me. Is anyone else having this problem and know of a fix?

I was having this problem when I upgraded 2.4.3... same symptoms... these blackouts would occure every 11 hours and to blocks of customers.... different blocks each time. These mini-stripe ICMP outages would last for 5-30min. Trying to ping an affected client radio and/or trying to access the managment page of the client radio would 'wake' them up and they'd start responding to ICMP. We never had any customers complain about the issue though, which makes me believe that customer traffic was somehow still getting through, even though ICMP to the SM's management IP would fail. The issues went away when we upgraded everything to 2.5, and moved over to GPS-sync, 75/25 ratio, and 2.5ms framing.

1 Like

Thanks for the reply. I upgraded my APs back to 2.5 and switched to 2.5 ms  framing and this time gave it 5-10 minutes for my client radios to reassociate with the AP and so far so good. I jumped the gun the first time when my client radios wouldn't connect back to my AP and switched back. 

Well I am still having the same issue after all.  If I downgrade my APs to 4.2.4 problem goes away. If I run 2.4.3 or 2.5 I keep having these random radio freeze moments. There has to be some explaination for it. I am using Mikrotik RB1100AHx2 routers and it happens most frequently to 1 tower but I have seen it on others as well. Randon radios stop responding even though they are still in session on the access point and they all drop at the same time from various APs on the same tower. They also still show up in the Mikrotik IP ARP table as they see the radios mac address but you can't ping the radio. If I deregister the frozen radios when they come back into session they are responding again. 2.5 firmware with 2.5 ms framing has given me my best throughput results but I can't run it with this intermittent issue. If I ignore the issue it takes an hour or so and they will start responding again.