Odd wifi issues after wireless health

[SFog]

We have a client with 10 MR42s on FW 25.13, out of nowhere they started having some very odd wifi issues, user will be randomly booted off the wifi to the point that if we reboot the computers they can contact the DC to auth. Wired works just fine and wireless was working fine about a month or two a go.

I have gone over the change logs to make sure nothing has been changed or updated, and it happens on all SSID on all APs on all brands of devices. We are seeing a bunch of Auth error in wireless health and a high 43% of failed connections. I have called Meraki and we found the wireless profile was set to a width of 80 so we cut that down to 20, but there was nothing in the change logs about this and I am sure no one would have changed it.

Has anyone else see odd wifi issues after wireless health when live? Or is this a known issue?  I will follow up if the width change helps.

[jdsilva]

Didn't wireless health go live like a year ago?

It didn't cause us any issues on a few thousand WiFi networks. In fact, it's been awesome to help troubleshoot quite a few issues. 

[SFog]

you are right, I was reading something that said it was new and saw beta but did not see the date on the article of almost a year ago.

[SFog]

Looks like I read a date Wong wireless health was not just released, it was released a year ago, so I am not sure what changed. but after the change from 80 to 20 the failed connections is down to 22%

[PhilipDAth]

I haven't had any issues.

You know the problem with shared RF spectrum is sometimes someone else does something that causes you a problem - even though you changed nothing.

It could be new cordless phones, RF microphones, cordless headsets, just about anything that uses RF.  Has anyone introduced new "anything" into the environment?

Don't rule out that the problem could be nothing to do with something you have changed.

It could even be someone malicious.

A channel width of 80Mhz is a tad hopefull.  I often like going for 40Mhz if the environment is not dense.

Does Wireless Sentry report any security issues?

Does the spectrum analyzer show all channels are good, or is part of the RF spectrum being chewed up by something?

And I hate to say it - but have you tried giving the APs a reboot?

[SFog]

RF

Things look good for the most part in the  RF spectrum and we also look for a spoofed SSID but not find anything. Lol yeah I did the ol reboot and things are still acting a fool. I have a call with meraki in the morning to take another look. It feels like something has changed but I was told from meraki that unlike the event logs (where they can see more information on the back end) the user see everything in the change logs and nothing has changed,.

[cta102]

Is the user completely kicked off (no wi-fi symbol) or is the user connected to the AP but has no network access and all other devices are fine?

I'm asking because I have had a few issues with our upstairs MR32 over the last 3 months or where the phone is connected to the AP but can't talk to anything (the dashboard says everything looks good.)

It started with the awful work phone I have, happened once with my decent phone and since the last firmware update happened a few times to two windows laptops (I have since downgraded the firmware to see if the issue occurs again.)

Rebooting the AP restores connectivity however I usually run on Beta firmware (it's at home so it's not a real issue.)

[SFog]

The users get completely kicked off, most of the time they can just rejoin or wait and windows will rejoin. This is happening with all laptops and ipads in the office. I was think it might be a wireless nic issue but we have at least 4 different types of winc  on 3 different brands of laptops and ipads. I thought it could be a windows issue too, but we have 2 different flavors of windows and mac os in the office.

[MMoss]

@cta102 

The MR32 has a defect that causes that kind of problem with certain subsets of WIFI cards. We just had some swapped out after one model of laptops we own had similar issues. 2.4Ghz was pretty bad, 5Ghz worked fine until the Beta tanked it and I rolled it back. Beta was an attempt to fix the problem you describe because I generally don't use them. 

[cta102]

@MMoss Thanks for that.

I's an issue I have only encountered since mid/late January and was causing issues on my work supplied Samung, as updates rolled out other devices started joining in my Google Pixel, sons iPhone and now two laptops (both the same model all using  5GHz.

Everything has behaved for a few days since rolling back

No doubt I will apply the next beta as it's the only way anybody is going to find out if the issues persist.

[SFog]

What firmware are you on now?

[cta102]

Back to MR 15.13 and so far there have been no issues in just under a week.

However it's really too early to tell as everything looked foe for a few days after the last beta release.

[MMoss]

@cta102 

That's cool. I'll be doing more testing with Meraki this evening with the Beta. It's for a different issue but I have a feeling this firmware will cause issues like it did last time and will overshadow and ruin the test. On the bright side they will be able to make the correction in the firmware with first hand captures on the consumer side so maybe it'll be corrected quicker. I'm always happy to help beta test personally as long as it doesn't get in the way of taking care of the people I'm responsible for 🙂

[SFog]

Update*

We have made a few changes, Meraki had some changes they recommend for the radio settings, lowering the power and channel width. We thought things were looking better due the the health of the network dropping down to 3% but it is now back up to 39% and we are still getting a ton of Auth errors. Meraki thought this was due to the laptops jumping from AP to AP, but changing the power settings has not changed anything.

[MMoss]

@SFog 

Factory Reset

Create a different SSID, Meraki once upon a time had a problem with SSID's in the 1st spot, but it may be resolved. They also suggested when working on an issue that I use WPA2 and not WPA 1 and 2.

Set Minimum Bitrate to 12Mbps to eliminate any old legacy equipment that may be causing problems. The long preamble and slow transfer rate mixed in with the half duplex nature of Wi-Fi can be a problem is someone is running around with legacy equipment

One more thing off the top of my head, and this is a longer shot than the previous 2. NAT Mode requires the device to re authenticate to each AP it moves to. This is because each AP works as a DHCP server and they don't share IP addresses between other AP's and themselves. 

Best guesses going through the post, good luck

[SFog]

Thank you for your help.

We do have a test SSID going right now what is WPA2 and I will report back how that works.

We have all SSIDs set to a min of 12, that fixed an issue we had when we first installed the APs.

only 1 of the 3 is using Meraki dhcp, but it in happening to all SSIDs.

[sebas]

any updates on this one ? We have .13 on MR42's and randomly access points stop forwarding traffic and have to be rebooted :S

[SFog]

As random as it started it stopped, I do think the wireless profile helped so make sure your APs did no get reset to the default profile of outside.