Configuring server redundancy for remote authentication is a best practice, especially for administrator access. When the primary RADIUS server is unreachable, FortiGate sends the authentication request to the secondary server.
Prerequisites: The administrator account must exist on all RADIUS servers, with the same credentials and attributes (group, Fortinet-Access-Profile VSA, if used). FortiGate must be registered as a RADIUS client on each server, with the correct shared secret. FortiGate must reach each server on UDP port 1812 (or the configured port).
Step 1: Configure the RADIUS server object:
GUI: User & Authentication -> RADIUS Servers -> Create New.
 CLI:
config user radius
edit "RADIUS-ADMIN"
set server "172.31.133.140"
set secret <primary-secret>
set secondary-server "10.133.5.120"
set secondary-secret <secondary-secret>
next
end
Step 2: Configure the user group.
GUI: User & Authentication -> User Groups -> Create New.
 CLI:
config user group
edit "RADIUS-ADMINS"
set member "RADIUS-ADMIN"
config match
edit 1
set server-name "RADIUS-ADMIN"
set group-name "super_admin"
next
end
next
end
Step 3: Configure the remote administrator. GUI: System -> Administrators -> Create New -> Administrator -> Type: Match a user on a remote server group.
 CLI:
config system admin
edit "test"
set remote-auth enable
set accprofile "super_admin"
set vdom "root"
set remote-group "RADIUS-ADMINS"
next
end
When an administrator logs in, authentication is handled by the primary server. If the primary server is down, the secondary server handles the authentication.
Verification: Run a packet capture on the RADIUS port:
diagnose sniffer packet any "(host 172.31.133.140 or host 10.133.5.120) and port 1812" 4
Sniffer 1: Successful authentication (through primary server).
diagnose sniffer packet any "(host 172.31.133.140 or host 10.133.5.120) and port 1812" 4
filters=[(host 172.31.133.140 or host 10.133.5.120) and port 1812]
14.321084 port2 out 172.31.133.115.15201 -> 172.31.133.140.1812: udp 111
14.337001 port2 in 172.31.133.140.1812 -> 172.31.133.115.15201: udp 20
Sniffer 2:Â Successful authentication (through secondary server) when primary is down.

diagnose sniffer packet any "(host 172.31.133.140 or host 10.133.5.120) and port 1812" 4
filters=[(host 172.31.133.140 or host 10.133.5.120) and port 1812]
16.139606 port5 out 10.133.5.115.1429 -> 10.133.5.120.1812: udp 111
16.149698 port5 in 10.133.5.120.1812 -> 10.133.5.115.1429: udp 20
From the output, traffic is going through the secondary one.
|