1131 AP's randomly disconnect

[tfetzer]

We are noticing that we are getting alarms and time stats showing several AP's disconnect and reconnect within a minute.  Is this normal behavior? I also ran a report for device Up Time and there are several that have an uptime less than a day.  We have 4 - 4404 controllers, software version 4.2.176.0, approx. 250 LAP1131AG-A-K9  AP's, WCS version 6.0.181.0 .  Thanks for any assistance.

[Leo Laohoo]

We have 4 - 4404 controllers, software version 4.2.176.0, approx. 250 LAP1131AG-A-K9  AP's, WCS version 6.0.181.0 .

If I recalled correctly, when I was using this WLC firmware version two years back, I remembered that our 1132 would regularly re-join the WLC.  After upgrading the firmware these "phenomenon" disappeared.

[George Stefanick]

Hello,

Good question. I actually just went through this myself. We have a few thousand APs and would get about 200 alarms a day of APs moving between controllers. So I started to dig into the "why" the last few weeks. I will be happy to share what I found ...

First, its important not to confuse APs that reboot and move controllers AND aps that hop around from controller to controller. If you have APs disconnected and rebooting.. this is likely a connectivity or hardware issue with the AP. You can determine this by looking at the AP uptime.

Click in wireless->select AP -> look on the right side.

UPTIME is how long the AP has been up.

Controller Associated Time is how long its been joined to the respected controller

What i found in my situation was that our WiSMs handle 150 APs per controller. We had like 500 APs assigned to controller (1) as the primary controller. Even though it can only handle 150. So once an AP moved off controller (1) for whatever reason there were 350 other aps looking to take that spot

We also found a lot of APs that did not have the primary, secondary or tertiary controllers set. We found a good portion of the aps that moved were configured like this ...

I hope this helps... If so, please rate helpful post.. And Happy Holidays to you !

[Surendra BG]

Nice One George!! 5+

[George Stefanick]

Well thank you Surendra. BTW, you have been a great help to me professionally... I always enjoy reading your replys.. Keep up the good work !

[Surendra BG]

Thanks George

[tfetzer]

Thanks for your great advise and explanation!

I looked at the AP uptime vs associated time:  Below is just one example.

Time Statistics
UP Time
Controller Associated Time
Controller Associated Latency

I am most concerned about the UP time since it looks like the AP's are disconnecting.

Controller 1 has 80 AP's; controller 2 has 67 AP's, controller 3 has 41, controller 4 has 42 AP's.  Controller's 1 & 2 are at one hospital; controllers 3 & 4 are at a different hospital.

We are running VoIP with cisco wireless phones for the nursing staff.  I am concerned that if the AP's are rebooting/disconnecting that the call is also dropping.

Here is an example of the alarms we see:

AP 'R26' disassociated from Controller 'x.x.x.x'.

802.11a interface of AP R26 is down: Controller 'x.x.x.x'      12/28/10 9:15:13 AM

802.11b/g interface of AP R26 is down: Controller 'x.x.x.x'      12/28/10 9:15:13 AM

Thanks again for all your help!

[Leo Laohoo]

Hmmm ... have you tried testing another firmware version?

[tfetzer]

No unfortunately I don't have anything to test a firmware upgrade.  Thanks though and I will keep that suggestion in mind.

[George Stefanick]

What I found in my situation ~~ I had a handful of APs that were just bad. One I determined was overheating and another that was upgraded from IOS to CAPWAP must have had issues during the upgrade.

These problem APs would reboot randomly.

I would see if there is any pattern and then monitor the switch ports and see whats up..

[tfetzer]

Thanks George!  We were also wondering that same thing.  Some of the AP's do

get very hot.  I'll post if I find a solution.

[elblanco_fr]

Hello,

I have similar problem :

few APs (10) randomly "disapear form wlc console" :

- i can ping those AP

- i can ssh to them.

but they are unusable

To solve this, i open en ssh session to the AP and reload it.

After it will works for many days (i don't know how many)

The 180 AP of my compaany are installed on more 70 sites

This problem appear on all the AP on 1 site  and on few AP on another site

On WLC all the AP have controller name controller IP address defined.

my configuration :

1 WLC4400 - 100AP

1 WLC5500 - 80AP

software version 6.0.199.0

[elblanco_fr]

i forgot some informations :

It looks like controller association problem :

A AP wich is anusable

UP Time
Controller Associated Time
Controller Association Latency

On the same place, an AP wich is avaible

UP Time
Controller Associated Time
Controller Association Latency

[elblanco_fr]

Finaly i found the solution, it's a bug n 6.0.199.x

https://supportforums.cisco.com/thread/2046449

sorry for "spamming" this thead, maybe it will help somebody.

[tfetzer]

Thanks for the information and link to the similar problem!  We are currently at WCS version

6.0.181.0 and not version 6.0.199.4.  So I'm wondering if the bug is also in our current version or only 6.0.199.