Cisco AIR-LAP1042N-N-K9 connecting to Cisco 2504 WLC drop outs

[jake.alamdar.1]

Hello,

I work in Australia and am currently haveing the following issue with my companies WLC and APs.

AP: AIR-LAP1042N-N-K9 (country code Australia)

WLC: 2504 (contry code Australia)

The light weight APs connect and join to the Cisco WLC and show this on the web GUI, but they randomly just dissconnect as say "not joined" then they will join again. Sometimes it will be stable for a while othere times it will happen quite frequently. I have attached some of the summarie errors from a few of the APs.

We have APs located in different parts of the world all connecting back to the same controller in our head office.

The error below is one that I see quite often, but cannot seem to find much information on the cause.

"number of message retransmission to the ap has reached maximum site"

If you need any mor einformation please do not hesitate to ask.

Any help would be greatly appreciated.

Kind Regards,

-Jake

[Leo Laohoo]

If the site does not have any local authentication server, then you should be in local mode.

[jake.alamdar.1]

I have enabled the link latency there were only two APs that were over the 300ms mark all the others were around 2-50ms Max.

[Scott Fella]

It's probably not that... is your AP's in local mode or FlexConnect?  Like Leo mentioned in his post, you should consider using FlexConnect mode and not local mode.  I will only use local mode when the AP's are located in the same building as the WLC or if they have like a OptiMAN or GigaMAN circuit.

Thanks,

Scott

Help out other by using the rating system and marking answered questions as "Answered"

[jake.alamdar.1]

Hi Scott,

All the APs are in flexconnect mode, maybe flex connect is misconfigure? I was just thinking of changing the APs on site to local mode.

[Scott Fella]

APs that are local where the wlc's are located should be left at local mode. Any other AP's at different building that goes thought your WAN should be FlexConnect. That's what I would do.

Sent from Cisco Technical Support iPhone App

[jake.alamdar.1]

I have shifted the APs that are located locally to the WLC to local mode. We just lost connectivity again, is there any way to increase the amount of time the WLC will wait for a heat beat from the APs?

[Leo Laohoo]

In your particular case, I don't believe enabling Link Latency will work.

It's happening now?

Do you have Telstra's GWIP service?

[jake.alamdar.1]

Yeah it is happening now. No we do not have Telstras GWIP Service. (im not entirely sure what that is).

[Leo Laohoo]

GWIP = Government-based Wideband IP.   It's Telstra's network for government.

[jake.alamdar.1]

Oh i see, no we dont have GWIP we are not a government organisation.

[Leo Laohoo]

Do you have QoS enabled in your network?

[jake.alamdar.1]

no i dont, what traffic would i give preference to in terms of the wifi? CAPWAP? LWAPP?

[Leo Laohoo]

LWAPP = CAPWAP

[maldehne]

when you do the required classification to pritize your traffic you should pay attention to the port numbers.

If you are using capwap , you should prioritize udp 5246 and 5247

if you are running old release you should prioritize UDP 12222 and 12223

[jake.alamdar.1]

Hey All,

Sorry I haven't replyed for sometime I was on leave. I managed to find the issue. What happened was a second cable was plugged into the WLC and back into the same switch, causing a loop. As soon as this was removed everything became satable.

I really appreciate all the help and advice on this issue. I learnt a lot.

Kind regards,

-Jake