topic Re: Access point auto discovery in Cisco Catalyst Center

Access point auto discovery

KevinR99 — Mon, 09 Oct 2023 09:48:03 GMT

I have an AP pool in my DNAC and as expected when I plug an AP into one of my access switches the port is auto configured to put the AP in the correct address pool. However, I have 9136 WAPs and they support LAG on their dual uplinks. Is there anything DNAC can do to push a LAG config out to the 2 ports the AP connects to or do I need to change the AP ports to LAG on the WLC and manually change the Edge ports ? The AP ports on the switch have a derived config which is removed if I remove the AP. If I were to manually set the ports in a LAG setup this would be a permanent setting meaning I cannot then remove the AP and revert the port automatically to a default setting.

Thanks, Kev.

Re: Access point auto discovery

Leo Laohoo — Mon, 09 Oct 2023 10:06:22 GMT

Try using both ports without configuring Etherchannel.

Re: Access point auto discovery

KevinR99 — Wed, 11 Oct 2023 07:27:52 GMT

So does than mean both ports are used even without etherchannel? If so why offer the option of etherchannel ?

Kev.

Re: Access point auto discovery

KevinR99 — Fri, 13 Oct 2023 12:30:48 GMT

The point here is to hopefully discover the AP on dual ports automatically and for the AP pool to be assigned. So then we don't need to go looking for the APs and manually assign their port. This works with an AP connected to a single port. The AP is identified via cdp and DNAC pushes out a config that can be seen with hte # show derived-config command. However, although the AP is also seen on another port via its second interface it never gets the derived config. When I shut the port that does have the config hoping the other port will then get the derived config it never does. So it seems DNAC cannot automate resiliency with dual AP connections.

DNAC is supposed to be about automation and it does work to an extend but not with a dual connected AP.

Kev.

Re: Access point auto discovery

Leo Laohoo — Fri, 13 Oct 2023 22:42:42 GMT

I am testing a 9136 dual-connected to a 9300 stack. The ports are static VLAN (non-Dot1X) and the ports, including the redundant PoE, works very well. I do not need to configure EtherChannel.

However, if dual-ports are Dot1X, it will not work. We have raised an Enhancement Request (CSCwc30701). This Enhancement Request will, possibly, NOT get any traction unless there is "executive support".