Aironet 1300 dropping every 36-48 hours, help please

[a_bullard]

Within the past month our 1300's have begun going down around every 36 hours. They are setup with a root bridge and 4 non-root bridges and are set to use the least congested frequency. I am interested in solving this problem as well as setting up the non-root bridges to act as APs as well as connecting some offices to our main network. Please help if you can! A

[Mr. Bash]

Can you describe what "going down" means?

[a_bullard]

the entire wireless drops. We have to do a hard reset on the main root bridge to get them back up. I was thinking is may be the DoS issue so I applied that fix last night and knock on wood, it was still up this morning.

[a_bullard]

Ok, so I was wrong. According to my monitor the traffic on the root bridge starts picking up at around 4:30 in the afternoon-surprising as most of the users are not even using at that time. Historically it build until the root bridge drops-can't ping it, web into it anything-dead-until I manually disconnect the power and plug it back in......

[andrew.brazier]

Faulty bridge? It it's that dead maybe a hardware problem? Can you console into it?

[a_bullard]

I have replaced the original with a spare and it still does the same thing. It is close to every 48 hours. I am getting this on the error log when they go down:

Critical RADIO crypto FIPS self test passed on interface Dot11Radio 0

17 Mar 1 00:00:03.953 Notification Interface Dot11Radio0, changed state to reset

18 Mar 1 00:00:03.607 Critical IOS crypto FIPS self test passed

The administration here is ready to throw the wireless out of the door, but I want to understand what is going on.

[a_bullard]

This is the error patern from an associated bridge:

Interface Dot11Radio0, changed state to up

2 Mar 10 15:27:15.265 Warning Interface Dot11Radio0, Associated To AP ClarendonHealth 0018.b92b.4190 [None WPA PSK]

3 Mar 10 15:27:15.060 Error Interface Dot11Radio0, changed state to down

4 Mar 10 15:27:15.059 Warning Interface Dot11Radio0, parent lost: Received deauthenticate (2) not valid

5 Mar 10 15:16:57.191 Notification Line protocol on Interface Dot11Radio0, changed state to up

6 Mar 10 15:16:56.191 Error Interface Dot11Radio0, changed state to up

7 Mar 10 15:16:56.190 Warning Interface Dot11Radio0, Associated To AP ClarendonHealth 0018.b92b.4190 [None WPA PSK]

8 Mar 10 15:16:56.003 Warning Interface Dot11Radio0, cannot associate: Associating

9 Mar 10 15:15:48.999 Warning Interface Dot11Radio0, cannot associate: No Response

10 Mar 10 15:15:39.446 Notification Line protocol on Interface Dot11Radio0, changed state to down

11 Mar 10 15:15:38.446 Error Interface Dot11Radio0, changed state to down

12 Mar 10 15:15:38.445 Warning Interface Dot11Radio0, parent lost: Too many retries

13 Mar 8 15:20:26.778 Notification Line protocol on Interface Dot11Radio0, changed state to up

14 Mar 8 15:20:25.778 Error Interface Dot11Radio0, changed state to up

15 Mar 8 15:20:25.777 Warning Interface Dot11Radio0, Associated To AP ClarendonHealth 0018.b92b.4190 [None WPA PSK]

16 Mar 8 15:20:25.193 Warning Interface Dot11Radio0, cannot associate: Rcvd response from 0018.b92b.4190 channel 9 3367

17 Mar 8 15:19:14.188 Warning Interface Dot11Radio0, cannot associate: No Response

18 Mar 8 15:19:04.550 Notification Line protocol on Interface Dot11Radio0, changed state to down

[andrew.brazier]

Very odd, not much information on that FIPS message on the web. Might be IOS version related maybe? Which version are they running? Also, when the root bridge goes down can you console into it or is it totally dead?

[a_bullard]

I have replaced the original with a spare and it still does the same thing. It is close to every 48 hours. I am getting this on the error log when they go down:

Critical RADIO crypto FIPS self test passed on interface Dot11Radio 0

17 Mar 1 00:00:03.953 Notification Interface Dot11Radio0, changed state to reset

18 Mar 1 00:00:03.607 Critical IOS crypto FIPS self test passed

The administration here is ready to throw the wireless out of the door, but I want to understand what is going on.

[a_bullard]

the root is totally dead. I have to physically go to it and cut the power to make it reboot.