04-06-2012 01:39 AM - edited 03-10-2019 06:58 PM
Hello,
This weekend we have upgraded the ios on quite a few switches on a larger site, the site is a mix of 2960 and 3560 switches and the previouse ios versions were 12.2.44 on most switches but some had an older 12.2.25.
On monday when we came into work we got a call that most of the ports on these switches were an amber color and most people could't use the network.
After some investigation we discovered that we had a problem with dot1x so for a quick solution we just removed it from the switches and restarted all the ports with no dot1x enabled, this solved the problem but we can't really figure out what exactly caused this to happen in the first place.
Our config looked like this:
aaa new-model
!
!
aaa group server radius etsdot1x
server xxx.xxx.xxx.xxx auth-port xxxx acct-port xxxx
server yyy.yyy.yyy.yyy auth-port yyyy acct-port yyyy
!
aaa authentication login default group tacacs+ local
aaa authentication dot1x default group etsdot1x
aaa authorization exec default group tacacs+ local
aaa authorization network default group etsdot1x
aaa accounting dot1x default start-stop group etsdot1x
and on the ports themselves:
interface FastEthernet0/1
switchport access vlan 20
switchport mode access
srr-queue bandwidth share 1 70 25 5
srr-queue bandwidth shape 30 0 0 0
priority-queue out
authentication port-control auto
dot1x pae authenticator
spanning-tree portfast
spanning-tree bpduguard enable
service-policy input PC-PORT-QOS-IN
if anyone could pitch any ideas as to why this might have happened ...
04-06-2012 10:26 AM
You would need to give more details. Were there authentication attempts on the Radius server /ACS ? or the switches were not sending anything at all ?
The dot1x system changed between those 2 releases. The commands starting with "dot1x xxxx" now start with "authentication ...".
Your port config is not bad per se, but it all depends on what you want to do. You do dot1x only right ?no mab ?
04-07-2012 02:45 AM
Unfortunately I didn't have the chance to look at the logs on the servers yet, they are being administrated by a different department and I couldn't get in touch with them yet. I suspect that the problem was something with reachability of the servers at the time as well but I just wanted to run the config by others in the meantime to make sure I didn't miss something else.
Also, no, we don't use MAB, that's the only dot1x related config we have.
So assuming the servers were reachble, is there any other factor that could prompt this reaction from the switches after an IOS upgrade?
04-08-2012 12:52 AM
Do you have any debugs? E.g. Debug radius authentication
Also check that 802.1x is globally enabled - I.e. dot1x system in global configuration mode.
Sent from Cisco Technical Support iPad App
04-09-2012 03:21 PM
as mentioned by justin - make sure you have enabled "dot1x system-auth-control" in global config.
Apart from this you stated that "switches were an amber color and most people could't use the network" this means that STP blocked these ports which has nothing to do with dot1x.
So make sure that during an IOS upgrade there is no stp topology change.
04-09-2012 11:15 PM
Not necessarily. If the station fails dot1x authentication, the port LED will also show amber, so it can be anything. Amber light means simply that port is blocked for whatever reason.
04-09-2012 11:42 PM
Well, I say that it was dot1x since it only worked after we removed the dot1x configuration from these ports and did a shutdown - no shutdown on them.
Also only access ports were affected, and they all have bpduguard on them so all the ports that were transmiting bpdus were working ok so I don't think spanning tree was involved.
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: