C9105AXIU-B int0 not pingable on subnet, webui not accessible

[Carlomd]

Hello all,

I have a C9105axi-b, which I know after reading around can be converted as a EWC ap, I followed a couple of youtube videos, I also did the day 0 setup, but somehow for the life of me, I can't reach or ping the INT0 of the AP, I'm coming from Aruba ap's so this is my first time using CIsco AP's. I have 2 9105axi-b, I will try the 2nd one if that works out of the box with the day 0 setup. I installed 17.15 fw when I converted to ewc. And have static IP on int0.

Here's my sh int br output sanitized, any ideas or pointers would be appreciated.

interface GigabitEthernet0
mac-address 0000.5e00.0101
ip dhcp client client-id GigabitEthernet0
ip dhcp client broadcast-flag clear
ip address 172.16.xx.x 255.255.255.0
no negotiation auto
!
ip default-gateway 172.16.xx.x
ip forward-protocol nd
ip tftp blocksize 8192
ip http server
ip http authentication local
ip http secure-server
ip http secure-trustpoint CISCO_IDEVID_SUDI
ip ssh bulk-mode 131072
ip route 0.0.0.0 0.0.0.0 GigabitEthernet0 250
ip dns server

[marce1000]

   - For testing I would have a go with this procedure (start from scratch)  :
          https://www.cisco.com/c/en/us/td/docs/wireless/controller/ewc/17-12/config-guide/ewc_cg_17_12/overview_of_the_controller.html#task_gs1_qzh_kpb

     Check if that makes a difference w.r.t the issues you are encountering
           (in this testing phase use DHCP for the interface too)

  M.

[Carlomd]

Hello Marce,

thanks for the reply, I have my 2nd 9105 AP testing right now, I converted to EWC, and then when it got to the day0 setup part, I chose dhcp, I'll see if that works.

[Carlomd]

Hello Marce,

ok it worked using dhcp, I can login to the webui, so if I wanted to change the int0 to static, I can just go ahead in the webui to do that? I see that int0 is in l3 mode.

[marce1000]

   - @Carlomd  I would advise to keep using DHCP at all times, if you want to try static, you will need to that
                       through the CLI in the running configuration (not with the GUI). But if you are connected over the
                       network and not through the console , then you get a chicken and egg situation (you may get disconnected).

                      You can verify addresses, by using the command show cdp neighbors detail
                      on the switch where the AP is connected to 

   M.

[Carlomd]

This is what I get for sh cdp nei det

Device ID: APA0A4.7F37.xxxx
Entry address(es):
IPv6 address: FE80::1526:B24:E94:7651 (link-local)
Platform: cisco C9105AXI-B, Capabilities: Router Trans-Bridge
Interface: GigabitEthernet1/0/48, Port ID (outgoing port): GigabitEthernet0
Holdtime : 140 sec

Version :
Cisco AP Software, ap1g8-k9w8 Version: 17.15.3.28
Technical Support: http://www.cisco.com/techsupport
Copyright (c) 2014-2015 by Cisco Systems, Inc.

advertisement version: 2
Peer Source MAC: a0a4.7f37.xxxx
Duplex: full
Management address(es):

[marce1000]

  - @Carlomd  Note that to verify addresses the show cdp neighbors detail command has to be executed on the
                       switch that services the network connection to the AP , not the AP itself>
                      If done in the above you only see a link-local address, so pointing to network issues or int0 not correctly
                      configured

  M.

[Carlomd]

Marce,

that is the output from my switch above.

[marce1000]

  - @Carlomd             >...that is the output from my switch above.
                        Ok, so if you don't see anything besides a link-local address then there is a problem,

 M.

[Carlomd]

As soon as I converted to static on cli, it lost connection and can't be pinged anymore, pretty strange, it kept the route 0.0.0.0 0.0.0.0 gig 0 250. Are these aps supposed to only work through dhcp, we have segmented vlans.

[marce1000]

  - @Carlomd    Are these aps supposed to only work through dhcp, we have segmented vlans.
                         For the time being I don't have an exact reason but it is better to keep using DHCP.
                         For your VLANs you can always configure an ip helper to point to the DHCP server for
                         that particular VLAN,

 M.

[Rich R]

I have it working fine, on 17.15.3, with static IP.

As I have pointed out on a few previous posts - the default EWC config "ip route 0.0.0.0 0.0.0.0 GigabitEthernet0" is in breach of Cisco's own Best Practice and guidelines.  Ethernet interfaces should never be used as a "next hop" on a static route - it's impossible for CEF to understand where the packets should be sent to - it needs a next hop IP which it can ARP for and then send the packets to the next hop MAC.  Interface can be used together with next hop IP to ensure the next hop is only reached through that interface but should never be used without a next hop IP.  Remove the existing static route and replace it with a valid static route.  Then save the config and reload the AP because CEF has usually got itself into a state at that point.
conf t
no ip route 0.0.0.0 0.0.0.0 GigabitEthernet0 250
ip route 0.0.0.0 0.0.0.0 GigabitEthernet0 172.16.xx.x
end
wr
reload

When you do "show ip rout" you should see:

Gateway of last resort is 172.16.xx.x to network 0.0.0.0
S* 0.0.0.0/0 [1/0] via 172.16.xx.x, GigabitEthernet0

And for my sanity please refer to the interface by the correct name GigabitEthernet0 not int0 ? <smile>

FYI: interface is specified as static route destination for things like serial and dialer interfaces or tunnels.

[Carlomd]

Rich, thanks for the detailed pointers, gotcha on the gig 0 rather than int0

so I followed your commands to remove the default route no route 0.0.0.0 0.0.0.0 GigabitEthernet0 250

instead with ip route 0.0.0.0 0.0.0.0 GigabitEthernet0 172.16.xx.x (my gw ip)

and now when I type sh ip rout I get this

Gateway of last resort is 172.16.xx.2 to network 0.0.0.0

S* 0.0.0.0/0 [1/0] via 172.16.xx.2, GigabitEthernet0
172.16.0.0/16 is variably subnetted, 2 subnets, 2 masks
C 172.16.xx.0/24 is directly connected, GigabitEthernet0
L 172.16.xx.8/32 is directly connected, GigabitEthernet0

And I cannot ping my gw ip, or any other ip's in the same subnet. The .8 address is gig 0 management int that was created when I ran the day0 config. I'm pretty stumped at this point.

[Rich R]

That looks good but did you save and reload after the change?
And are you sure you removed the old static route - what you've typed above is missing the ip from "no ip route"?
What does your switchport configuration look like?

[Carlomd]

I did reload after wr mem, for some strange reason it keeps putting back the default route after issuing a 

no ip route 0.0.0.0 0.0.0.0 GigabitEthernet0 250

it would be there again after reloading, so I have a catalyst 9200 sw and I read I'm supposed to use a tru on it, so I have this as my trunk where the 9105 is connected.

interface GigabitEthernet1/0/48
description testwifi
switchport trunk native vlan 60
switchport mode trunk

and for the "sh cdp nei det" I get this when I run it

Device ID: APA0A4.7F37.E4AC
Entry address(es):
IPv6 address: FE80::9C:4875:6E9:4C4 (link-local)
Platform: cisco C9105AXI-B, Capabilities: Router Trans-Bridge
Interface: GigabitEthernet1/0/48, Port ID (outgoing port): GigabitEthernet0
Holdtime : 166 sec

Version :
Cisco AP Software, ap1g8-k9w8 Version: 17.15.3.28
Technical Support: http://www.cisco.com/techsupport
Copyright (c) 2014-2015 by Cisco Systems, Inc.

advertisement version: 2
Peer Source MAC: a0a4.7f37.e4ac
Power drawn: 11.682 Watts
Management address(es):

someone else said if it only shows link local, there might be something wrong.