WLC 5508 with LAP-1142n - Several Errors

[b.petronio]

Hello all,

I had installed a WLC 5508 with 7 LAP 1142n and 2 converted AP 1131abg.

I am seeing some errors relating 2 issues.

1st- One particular AP 1142 is disassociating and reseting the radios.

Thu Oct 28 11:50:49 2010	AP's Interface:0(802.11b)   Operation State Up: Base Radio MAC:e8:04:62:23:ac:e0 Cause=Radio interface   reset. Status:NA
Thu Oct 28 11:50:49 2010	AP's Interface:0(802.11b)   Operation State Down: Base Radio MAC:e8:04:62:23:ac:e0 Cause=Radio interface   reset. Status:NA
Thu Oct 28 11:50:49 2010	AP's Interface:1(802.11a)   Operation State Up: Base Radio MAC:e8:04:62:23:ac:e0 Cause=Radio interface   reset. Status:NA
Thu Oct 28 11:50:49 2010	AP's Interface:1(802.11a)   Operation State Down: Base Radio MAC:e8:04:62:23:ac:e0 Cause=Radio interface   reset. Status:NA
Thu Oct 28 11:50:46 2010	AP's Interface:1(802.11a) Operation   State Up: Base Radio MAC:e8:04:62:23:ac:e0 Cause=Radio reset due to Init.   Status:NA
Thu Oct 28 11:50:46 2010	AP's Interface:0(802.11b)   Operation State Up: Base Radio MAC:e8:04:62:23:ac:e0 Cause=Radio reset due to   Init. Status:NA
Thu Oct 28 11:50:46 2010	AP 'AP3', MAC:   e8:04:62:23:ac:e0 disassociated previously due to AP Reset. Uptime: 1 days,   10 h 24 m 23 s . Last reset reason: operator changed 11g mode.
Thu Oct 28 11:50:35 2010	AP Disassociated. Base Radio   MAC:e8:04:62:23:ac:e0
Thu Oct 28 11:50:35 2010	AP's Interface:1(802.11a)   Operation State Down: Base Radio MAC:e8:04:62:23:ac:e0 Cause=New Discovery Status:NA
Thu Oct 28 11:50:35 2010	AP's Interface:0(802.11b) Operation   State Down: Base Radio MAC:e8:04:62:23:ac:e0 Cause=New Discovery Status:NA

I had some search, and the new discovery cause, might be that the AP didnt know what WLC do associate, in a multi-controller environment. This is not the case. I only have one WLC in the same management vlan.

2st-The Radius server is beeing related in the logs as been deactivated. I raise the server time-out on Radius configuration option, but it still continues to do it.

Thu Oct 28 10:24:41 2010	RADIUS server 10.67.128.36:1812 deactivated in global list
Thu Oct 28 10:24:41 2010	RADIUS server 10.67.128.36:1812 failed to respond to request (ID 172)   for client e8:06:88:51:c0:2b / user 'unknown'

Is this meaning the WLC stop sending request to the Radius Server ? We dont have BackUp Radius.

As far as i know, its always the same mac-address client that is associated to that error, maybe a iphone.

I had so many clients in that SSID and they are all working good.

The Radius server is a NPS from windows Server 2008, and the client says that the medium response time is 0,02 sec, so im wondering why the controller is not getting response from Radius for a particular client?! My client also says, that didnt found any log related to that mac-address client ... what is weird...

WLC with last software available 7.0.164

Hope some one help me here.

Best Regards,

Bruno Petrónio

[Leo Laohoo]

What country domain did you enable on the WLC and what is the AP's complete model number?

[b.petronio]

Hi,

country domain:

config country PT

AP's complete model number:

AP 1 - 6, 9  = AIR-LAP1142N-E-K9

AP 7 - 8 = AIR-LAP1131AG-E-K9

I had run WLC Analyzer and the following warnings:

- After read the meaning i still cant understand what is missing

Type	Object	Warning
Controller	CGPTLISBWLC01	30002,General: Controller with APs with AP-Group in use

30002, General: Controller with APs with AP-Group in use:  Just a notification, so VLAN mappings can be validated, taking them  into account for troubleshooting. Also this affects  AppleTalk/IPv6/multicast so it is important to know.

- For all APs.

20012,AP: Empty primary controller. It is  recommended, to have a primary controller name configured, for  better/more predictive AP join process. This is not mandatory

20017,AP: Syslog messages are sent to broadcast  address, if there are errors reported by many APs, and there are too  many APs per vlan, this can cause broadcast storms. For best practices,  it is better to configure to individual server

I cant see relation to the messages i reported before.

Best Regards,

Bruno Petrónio

[Scott Fella]

Have you tried to swap out that 1142 to a different location with one that isn't giving you issues?  Just to test:)

For your radius issue, you can try to increase your Radius timeout:

config radius auth retransmit-timeout 1 5

1 = which radius server, since you only have 1, #1 is probably what you need to use

5 = # of seconds

[b.petronio]

Thank You Scott,

The Radius server timeout is indeed with 30 sec.

Initially was 2, and we change it for the maximum possible.

Regarding the 1142, are you sugesting that it could be the place were the AP is, and not the AP itself that is causing that issues ? Interferences.. etc ?

Best Regards,

Bruno Petrónio

[Scott Fella]

No... If you take an AP that is working fine and move that to the location where you have that issue that will eliminate issues with the AP and might point to cabling or swicthport or something else. If you move the AP to a different location and you still have the same issue, then its probably the AP.

Posted from my mobile device.

[b.petronio]

Thanks Scott,

I understand what you are mentioning, and i really didnt do it yet.

I realize that the primary controller was not configured on the Wireless –> All APs –> High Availability tab, and did it only to the AP that is taking this beahviour.

Is this mandatory for a 1 controller only ?

No mather what the manual say, after that the AP is rebooting 2 mins in 2 mins... with the same kind of messages.

The interface on the switch is getting a few input errors and the same numbers of crc... but are so few...

Next step ... i will change it to another one's place/pathing cable.

Regarding the Radius messages... any ideas ?

I'm already on 30 sec's of server timeout.

Best Regards,

Bruno Petrónio

[b.petronio]

Hello,

I will swap the AP position with another one in the network to make the cotton test .

Nevertheless i see the messages attacthed 2 files with log messages from the AP with the several errors reported.

This is the main error reported when it disassociates from the controller.

(%CAPWAP-3-ERRORLOG: Retransmission count for packet exceeded max(CAPWAP_WTP_EVENT_REQUEST., 5))

Its not always, but several times before show that event, i see logs from GigaEthernet interfece going down and up .

I hope someone could help me on this.

Best Regards,

Bruno Petrónio

[Scott Fella]

Well since you did swap the AP to a good know cable and switchport, I would say the AP is bad.  You could always RMA that AP or if you have time, convert that to an IOS AP and then back to a LAP and see if that fixes your issue.

[b.petronio]

Scott,

The AP stops broadcasting that messages, but when we realize, it connects to the switch at 100Mbps. lol, even if the switch has auto/auto on speed/duplex configuration.

***

I had a new AP that is saying other message. Can anyone advise me ?

Tue Nov 16 11:07:24 2010

AP 'AP2', MAC: e8:04:62:0a:c7:a0   disassociated previously due to AP Crash. Uptime: 0 days, 00 h 00 m   59 s . Last reset reason: .

Best Regards,

Bruno Petrónio

[Surendra BG]

Hi Bruno,

The latest msg means the Ap has crashed and the crashino file or the core dump will be generated and you can see those in the management >> Tech support >> AP core dump or >> AP Crash log.

please open up a SR with us to get the Root Cause Analysis...

Regards
Surendra