Controllar associated time

[jain.manish94]

Team there are 2 to 3 access point at the site . The problem is that controller associated time is changing after few minutes so all users are disconnecting.  Plz let me know what is the problem.

[Leo Laohoo]

Is this FLEX?

[jain.manish94]

Yes.

[jain.manish94]

All are flex.

[Jurgens L]

Are you having this issue at the one site only and other sites are ok?
If so, check the latency of your WAN connection to the branch.
On the link below you will find table 7-3 which will show the bandwidth and latency requirements for flexconnect AP's.
https://www.cisco.com/c/en/us/td/docs/wireless/controller/8-5/Enterprise-Mobility-8-5-Design-Guide/Enterprise_Mobility_8-5_Deployment_Guide/ch7_HREA.html#pgfId-1078488
If latency is a problem you will need to look at enforcing a QoS policy to prioritise UDP port 5246, this is the control traffic that flows between the AP and the WLC ensuring the AP is associated to the WLC.


<<< Please help the community by marking useful posts helpful, or accept as a solution if it resolved your issue >>>

[jain.manish94]

U are talking about controller association latency for the ap that is 11 seconds?

[patoberli]

No, that is how long the AP needed to connect to the WLC the last time it booted.
You can for example test with ping ip-address-of-ap from the WLC if you have the issue. I'm not sure if you somewhere can see the CAPWAP latency.

[jain.manish94]

It is big issue for me because in off hours the particular access point controller associated time remdonly changing. Some time after 30 min some time after 40 min and after that all users are disconnecting from that ap and connecting again. 

[patoberli]

If you ping now an AP from the WLC, what latency do you get?
Do you monitor the internet line between the AP and WLC and is that line ever >80% loaded?
If yes, do you have QoS in place?

[Jurgens L]

The latency for a packet to travel from the branch site to where the WLC is hosted and back (round trip). If you monitor your WAN links throughput and latency you should see if there is any congestion or high latency over a given period.

If you don't have any monitoring in place, you will need to run continuous pings, either from the branch to the WLC or vice versa and keep an eye on it, if you see your latency exceeds over 300ms this will explain why the AP dissociate from the WLC.

If you don't have any CRC errors on your WAN links, the problem will be most likely be congestion and there is more than likely no QoS configured for the UDP port 5246 to be prioritized over AP control traffic over the WAN.

[jain.manish94]

Team still same issue controller associated time getting zero . This is with same switch and same port when I am connecting any access point same issue but the same access point is working in other place without any issue . What is the problem. ?

[patoberli]

That still indicates a cabling or power issue at that location. Please validate the cable and if it supplies enough power.

[jain.manish94]

Cable has been changed still same issue

[patoberli]

Is it a manageable switch to which your AP is attached to?

If yes, can you check its logfiles for any indications?

[Sathiyanarayanan Ravindran]

Hi,

Go to Advanced setting inside the access point and try to change the AP Retransmit Config Parameters from default values to Values mentioned as per below screenshot on one Access Point and check the performance.