Upgrade WLC to 8.5.120 - advice please

[Jay_F]

Hi,

Thanks for those who had helped me through the upgrade of our WLC's from 8.0.133 to 8.5.120. I upgraded the HA controller to the interim 8.3.140 and then to 8.5.120. I moved the AP's off the primary to the HA and then upgraded the primary to 8.5.120. However, and i'm not entirely sure it is related but our clients have since been unable to access The Cloud. i had wondered if it had anything to do with the anchor controller now being on and older version to the primary and HA ? - i wouldnt think its likely as the other SSID's we have users are able to connect. Only seem to affect access to The Cloud. The AP's being used are Cisco Aironet 2700 series. Any ideas ?

Thanks in advance.

[Leo Laohoo]

---

@Jay_F wrote:

However, and i'm not entirely sure it is related but our clients have since been unable to access The Cloud

---

"The Cloud"?  Care to explain what this is?  

[Jay_F]

Its public wifi that we allow and not working since the HA and primary controller has been upgraded to version 8.5.120. 

[Leo Laohoo]

Are the clients, connecting to "The Cloud", getting a valid IP address or not?

[Jay_F]

No they all get a 0.0.0.0 address shown on the wlc. Cant figure it out as nothing changed in configuration of public wifi.

[Leo Laohoo]

---

@Jay_F wrote:
No they all get a 0.0.0.0 address shown on the wlc. 

---

Check for Allowed VLANs and see if the public WiFi Dynamic Interface is allowed through the link.

[Jay_F]

I wonder whether it would be possible to downgrade the HA controller back to its original 8.0.133, move my test AP from primary over to HA after the downgrade just to confirm it is firmware related ? that should be possible right.

[Leo Laohoo]

I don't think this is firmware related.

[Jay_F]

looking at the mobility group on the anchor controller i can see that the path to this particular primary controller is showing 'control path down' - what could this be, nothing changed in our configuration other than the primary controller having the latest firmware installed.

[Jay_F]

Could anyone advise why the control path status would be set to down after upgrading the primary and ha controller to 8.5.120 - I am pretty sure the upgrade has impacted this particular sites public wifi. I haven't upgraded the other sites that use this anchor controller and their public wifi is working fine. Anyone have any tips of what to look for ?

[patoberli]

Make sure the MAC addresses are still correct in the mobility configuration. Also check if you now have "new-mobility" enabled, as far as I remember that was added some when in the 8.x versions.
Can you ping the IP addresses which you use for mobility of the other controllers from the CLI?

[Jay_F]

ok mac address all ok. enable new mobility ticked. 

I logged into the cli of the ha controller and pinged the anchor controller and had replies. I logged into the cli of the anchor controller, pinged the ha controller and had no replies. Any ideas? only happened since the upgrade of the primary and ha controller.

[patoberli]

I think I remember something about having needed to change a setting when I upgraded.

In any case, I don't use new mobility, only the old one.

What I did though, I enabled Multicast Mobility because of another problem in mobility.

Can you ping the HA mobility IP address from another client in the same subnet or from the router?



Oh, and verify your MAC address again, on the HA controller it's listed under Controller - Mobility Management - Mobility Configuration - Mobility Mac Address. You have to use that one in the mobility configuration for both controllers.

[Jay_F]

Checked the mac's again and all ok. Tried pinging the HA mobility ip address from another client on the same subnet as HA and had replies. Can ping the anchor controller from the HA, but cant ping the HA controller from the anchor controller. Status is still set to 'control path down'.

[patoberli]

I don't have the anchor controller configuration, only a HA and a standalone WLC here. From that standalone (running 8.5mr3 beta) I can ping the HA which is running 8.2.161.0.

Not sure if you are hitting a bug here.

I assume, because you write anchor controller, they are not in the same subnet?

If yes, then I sadly can't help much more.

Maybe you could try to some debug commands, they start with 'debug mobility'.

But I think the main problem is the missing reachability from the anchor to the HA. As far as I remember, ping must work.