Symptoms
We deploy an SSO pair of 5520's in our DC and DR sites running 8.3.143.0 code.
Our Enterprise WLAN had been somewhat reliable until several months back. Suddenly users began creating service tickets complaining that one or sometimes both of our production SSID's would deny them connectivity.
As I gathered symptoms while troubleshooting this I found that when this failure did manifest itself, that no parallel could be drawn with respect to device and/or model of access point as being the culprit.
Diagnosis
Back on 11/21/2020 I remembered reading a post form "sagarshaha" stating that he had trouble (specifically with Apple devices) and he had played with (FT) Fast Transition settings in an attempt to resolve his intermittent connection issues. But never saw a follow-up as to if he had seen confirmed success in his trials.
Our intermittent behavior was different than his, in that ours was exhibited on various device models and OS types (Win10, Android, Apple), but I thought if I created an alternate SSID for this test (so as not upset production) and acquire a sampling of wireless behavior (absent FT) in a very granular test group, I could then have users use this alternate SSID at precisely the moment they were experiencing the "failure to connect" type of symptom.
Intermittent issues are the worst, Right??? I mean they're hard to troubleshoot because they may not persist long enough for you to gather any confirmation of your fix. And also hard to know if the issue has been resolved, because it takes so long to determine if your fix was really instrumental in the resolution.
In any event, over time I began seeing good results in testing with the new SSID without Fast Transition, and later created a deployment event to employ this condition on the Prod SSID's.
Since this time, I have not had a single "failed to connect" symptom reported from a client device.
I cannot comment as to why this symptom was so intermittent, it was very random all over the country for us, some sites never experienced the condition, others 3 times a week, etc.
At a later release I see myself re-deploying (FT), possibly 8.5.X (or higher), but for now supporting legacy APs I cannot upgrade, so I knew I had to find a suitable alternate solution.
Solution
Disable (FT) on all production SSID's.
