10.2r4 firmware causing AP650's to lose connectivity?

[KevinP1]

Okay, I don’t know what’s happening here but this isn’t looking pretty. All of our AP650’s are losing connectivity multiple times throughout the day. 

Look at this! Some AP’s will only do this once or twice and others are doing it often, like this ? There’s nothing indicating any type of error or issue. We didn’t have this problem on firmware 10.09rb.

I have a case open with GTAC, but are we the only ones seeing this issue?

[Hans_SD44]

I mean, 10.09rb was stable for months but now even that version shows the same reporting after rolling back from 10.2r4. 

[SamPirok]

Thanks for clarifying, but that shouldnt make a difference here. They just wanted to see if 10.2r3 was stable, regardless of what version you moved from. I appreciate the thought though! 

[Hans_SD44]

@Sam Pirok 

Oops, bit braindead this morning, that AP wasn’t actually a rollback, as I had moved it from 10.2r4 to 10.0r9b previously. So, that graph shows a move from 10.09rb to 10.2r3. Not sure that really makes a difference, but just FYI so all the info is correct.

I will find one on 10.2r4 to rollback to 10.2r3 and give it 8+ hours and see how it looks.

[SamPirok]

Hey guys, good morning. Thank you @javabomberman for testing that for us, I will pass your results and notes on to the engineering team working on this. 

@kevin.piazza , I haven’t heard anything since the suggestion to try 10.2r3 yesterday, but it looks like that likely won’t help here based on javabomberman’s results. I will start pinging that team as soon as I see them sign on for the day and I’ll let you all know what they have to say. 

[Hans_SD44]

@Sam Pirok 

Here’s the device I rolled back to 10.2r3 last night:

Looks like no change on the memory dump reporting.

As a note, I still haven’t had any significant issues with the vast majority of my fleet on 10.2r4. I see the same reporting above on all of them, but I’m just not sure what’s actually happening at those points and it’s clearly so random that I’m not sure how to actively monitor for that effectively.