Network Management

I am seeing this error across all of my APs. 

I looked at congestion, however, all have less than 30 users. Most have about 10, with only one having 30. I also looked at noise and interference and only a handful of our 41 APs have noise issues that exceed the threshold of -90 dbm. Those APs that do report noise issues are between -85 to -89 dbm. 

We have several remote sites with APs. Those sites are connected via MPLS with Citrix SD-WAN. The Airwave sits in our DC and manages all APs local and remote

The impetus for researching this is that users are complaining that they are being disconnected from both guest and employee wifi several times throughout the day. Specifically, their RDP sessions are dropping.

Our employee wifi is secured via EAP-TLS.

I am not sure where to go at this point. The fact that I am seeing it across all APs leads me to believe it may be something at the route/switch level and not necessarily a wireless issue.

I haven't looked into yet, but could this be a mismatch in MTU settings somewhere?

Also, pings to/from a wireless client are odd. see attached screen shot.

Thanks for any insight and help!

Brian

Hi Brian, 

I am assuming you are seen these Frame Errors/sec on the local APs too, right? So we can remove the MPLS Citrix SD-WAN from the equation.

1. Monitor uptime of APs. Are they rebooting ?
2. Monitor "tot-t" value in the show ap bss-table <ap-name> . This will help determine the radio up time for that AP.

If the AP itself is rebooting, users may lose few packets ( causing RDP to drop ) while reconnecting to the next  available AP.

The AP may stay up but Radio resets may be happening? causing the clients to reconnect.

Try to stablish a pattern for RDP disconnects, if users are diconnecting at same time on different APs & locations; make sure you are not hitting something like : https://support.microsoft.com/en-us/help/2083411/remote-desktop-sessions-may-be-disconnected-during-group-policy-update

DSP

Thank you for the reply!

The RDP disconnects look to be related GPO updates. That was a new one to me. Learn something new everything day!

As far as the ping/frame retrys: I just got off the phone with TAC and given how sporadic the spikes/drops are they didn't want to make any changes. Their suggestion was to enable 80mhz support for channel width if  things become worse. So I will be monitoring and proceed from there.

No.  Enabling 80mhz channels would not help with low bandwidth traffic like RDP.  It will make your channels wider, but allow for fewer channels and could cause more interference, as a result.  Enabling 80 mhz channels should not be something to do in this situation.

We determined that the RDP issue is completly unrelated to wireless.

Enabling 80mhz channels was Aruba TAC's suggestion to deal with the sporadic latency spikes and dropped packets.

I understand.

Enabling 80mhz channels is also not a solution to latency spikes.  Enabling 80 mhz channels by default only leaves you with two channels which could cause alot of cochannel interference.  It could make your RF problem worse.

You would enable 80mhz channels:

- When you are sure the majority of devices support 80mhz channels and possibly DFS channels

- When you have low density of access points

- When you have the situations above and you want to support high raw throughput.

The 80mhz channels didn't make a whole lotta sense to me and was curious as to why TAC would suggest it.

Any suggestions?

Brian B

If you fixed your RDP issue, there is nothing to fix, is there?  Periodically high bandwidth and increasing errors are par for the course with running a WLAN.  You just need to design to make sure that your network can handle it when it happens.