Failover Radius issue with c2950G

[gilou_1973]

Hello,

I am using a DOT1X authentication based on two IAS Windows Servers.

My access switches are C2950 and I want to work with multiple Radius server for high availibility purposes.

But when my primary IAS server falls down (is declared dead when I debug radius messages), my switch try to connect with the second one and it doesn't work.

I've got the following messages :

"Sep 12 10:22:57.412 UTC: RADIUS: Re-signed packet (key: ciscomgt; rctx: 0x80DA94

18)

Sep 12 10:22:57.412 UTC: RADIUS: Trying next server (10.4.20.112:1812,1813) for

id157

Sep 12 10:22:57.412 UTC: RADIUS: Fail-over denied to (10.4.20.112:1812,1813) for

id157

Sep 12 10:22:57.412 UTC: RADIUS: No response for id 157

It's always in "fail over state" with the second one.

With I try separetely, the two servers works correctly. When I try with a C2970 plateform it works too. Is there special recommendations with the C2950?

Thanks for your helps

[Jagdeep Gambhir · Accepted solution]

Hi Gilou,

Please add the following command and if possible try the failover again

radius-server deadtime 1

Let me know how it goes

Regards,

~JG

Do rate helpful posts

[gilou_1973]

Hi jgambhir

Thanks for the advice. The problem is resolved.

The requests are now sent and processed by the backup IAS server.

By