If at boot time you have configured the following:
Router(config)# aaa new-model
Router(config)# radius-server retry method reorder
Router(config)# radius-server retransmit 0
Router(config)# radius-server transaction max-tries 6
Router(config)# radius-server host 10.2.3.4
Router(config)# radius-server host 10.5.6.7
and both servers are down, but not yet marked dead, for the first transaction you would see the transmissions as follows:
10.2.3.4
10.5.6.7
10.2.3.4
10.5.6.7
10.2.3.4
10.5.6.7
If you configure the reorder as follows:
Router(config)# aaa new-model
Router(config)# radius-server retry method reorder
Router(config)# radius-server retransmit 1
Router(config)# radius-server transaction max-tries 3
Router(config)# radius-server host 10.2.3.4
Router(config)# radius-server host 10.4.5.6
and both RADIUS servers are not responding to RADIUS packets but are not yet marked dead (as after the NAS boots), the transmissions
for the first transaction are as follows:
10.2.3.4
10.2.3.4
10.4.5.6
Subsequent transactions may be transmitted according to a different pattern. The transmissions depend on whether the criteria
for marking one (or both) servers as dead have been met, and as per the server flagging pattern already described.
If you configure the reorder as follows:
Router(config)# aaa new-model
Router(config)# radius-server retry method reorder
Router(config)# radius-server retransmit 1
Router(config)# radius-server max-tries-per-transaction 8
Router(config)# radius-server host 10.1.1.1
Router(config)# radius-server host 10.2.2.2
Router(config)# radius-server host 10.3.3.3
Router(config)# radius-server timeout 3
And the RADIUS server 10.1.1.1 is not responding to RADIUS packets but is not yet marked as dead, and the remaining two RADIUS
servers are live, you see the following:
For the first transaction:
10.1.1.1
10.1.1.1
10.2.2.2
For any additional transaction initiated for any transmissions before the server is marked as dead:
10.1.1.1
10.1.1.1
10.2.2.2
For transactions initiated thereafter:
10.2.2.2
If servers 10.2.2.2 and 10.3.3.3 then go down as well, you see the following transmissions until servers 10.2.2.2 and 10.3.3.3
meet the criteria for being marked as dead:
10.2.2.2
10.2.2.2
10.3.3.3
10.3.3.3
10.1.1.1
10.1.1.1
10.2.2.2
10.2.2.2
The above is followed by the failure of the transmission and by the next method in the method list being used (if any).
If servers 10.2.2.2 and 10.3.3.3 go down but server 10.1.1.1 comes up at the same time, you see the following:
10.2.2.2
10.2.2.2
10.3.3.3
10.3.3.3
10.1.1.1
When servers 10.2.2.2 and 10.3.3.3 are then marked as dead, you see the following:
10.1.1.1
Feedback