10-16-2008 11:42 AM - edited 07-03-2021 04:38 PM
We have access points spontaneously jumping between 2 controllers. When this happens, the WLAN associations on the radios are lost. Each time an AP jumps to the other controller, the WLAN is re-assigned to the radio and the configuration is saved. This is highly problematic because the WLAN that is lost is our voice WLAN and calls are being droppped. Can anyone explain why this is happening?
We are running software version 5.1.151.0 on the WiSMs and using 1252 (a/b/g/n) access points.
Any help would be greatly appreciated.
Ryan
10-16-2008 11:54 AM
Ryan,
An ap will only join another controller if it has lost its communication with the primary wlc. What I would do is disable ap fallback so that your ap's are not bouncing around. this will stabalize your netwrok for now, but you must troubleshoot why the LAP's are loosing connectivity to their primary. Could be an infrastructure issue or a wlc configuration issue... jsut review everything again.
10-16-2008 12:06 PM
Can you advise on things to look for? I don't see anything in the WLC logs or 6509 logs
10-16-2008 12:17 PM
What have you observed... do you see the ap's trying to get back to the primary or are they actually joining the primary again and then after x amount of minutes, hours, days, the ap's fails to the secondary? Is it all the ap's on a wlc or just some. Let me know more detail, because it can be many things.
10-16-2008 06:55 PM
What I would also look at at the uptime of the AP... this can tell you if the ap lost power or went through the discovery process. Also note the association time of the ap.
10-17-2008 05:38 AM
I would just like to add my thoughts to this post. I have noticed a few oddities with the AP fallback on my networks. I've determined that if you have WLC's running different code revisions, the AP's will have to download each particular mini ios before it can register. I have been able to stabilize the AP's falling back to their correct controller (eventually) by doing the following.
1. Make sure a Primary controller name is set in the AP config. It must match the system name of the WLC.
2. Try to have each WLC on the same code version. This will prevent the extra code download when they fallback.
3. If you have a WCS Server, do the Configure--> Controllers --> Refresh Config from Controller, and then choose the "delete" option. This will sync the info you programmed into the WLC into the WCS server's database.
Hope this helps!
Craig
10-17-2008 09:25 AM
What you want to look for are heartbeat timeouts. If you see heartbeat timeout errors then the AP is losing communications with the controller. This happens often when layer 3 lwapp control packets are blocked or not forwarded.
10-20-2008 07:17 AM
The APs get the controller information via DNS when booting up. So far, only a few APs are having this issue and it occurs spontaneously. The only common thread is that all of the APs experiencing this are the new 1252 a/b/g/n APs and are isolated in one closet.
The APs get controller information via DNS on bootup. Sometimes the APs join the controller specified in DNS and other times they join the second controller.
The AP power injectors are plugged into a UPS (with very little load). The WLC and 6509 logs do not show anything indicating a problem.
That's probably not very helpful, but that's all I have at the moment.
10-20-2008 07:18 AM
The biggest problem is that the WLAN associations are being lost. Even if the APs "fall back" to a controller they were previously associated to, wouldn't they maintain the WLAN associations on the radio?
10-20-2008 07:53 AM
You might call TAC and have them publish 4.2.130.6 for you. It is only available from TAC but has several 1252 fixes in it.
10-20-2008 08:08 AM
That'll be a problem because we're running Spectralink VoIP handsets that require WLC v5 software.
10-20-2008 08:24 AM
Yep. You got a problem. LOL
12-23-2008 12:05 PM
Which version of Spectralink handsets do you have? We've been running controller code 4.2.130 due to Vocera and haven't had an issue with Spectralink.
01-05-2009 07:48 AM
We're using the 8030 handsets.
01-05-2009 01:49 PM
Had kinda a similar problem...Aps would not spontaneously move from controller to controller but some would ignore the primary controller config. and move to another controller (eventually). Although we had controller for specific APs and do not want them to move, I configured all the mobility stuff (RF names) adding controller mac addresses and all the stuff. Once I was done, I restarted both controllers, and the APs found their primary controllers and have not moved since (about 5 weeks now).
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: