Add weight setting on each link health monitor server

Prior to FortiOS 7.0.1, the link health monitor is determined to be dead when all servers are unreachable. Starting in 7.0.1, the link health monitor can configure multiple servers and allow each server to have its own weight setting. When the link health monitor is down, it will trigger static route updates and cascade interface updates if the weight of all dead servers exceeds the monitor's fail weight threshold.

config system link-monitor
    edit <name>
        set srcintf <interface>
        set server-config {default | individual}
        set fail-weight <integer>
        config server-list
            edit <id>
                set dst <address>
                set weight <integer>
            next
        end
    next
end

server-config	Set the server configuration mode: default: all servers share the same attributes. individual: some attributes can be specified for individual servers.
fail-weight <integer>	Threshold weight to trigger link failure alert (0 - 255, default = 0).
server-list	Configure the servers to be monitored by the link monitor.
dst <address>	Enter the IP address of the server to be monitored.
weight <integer>	Weight of the monitor to this destination (0 - 255, default = 0).

Examples

In the following topology, there are two detect servers that connect to the FortiGate through a router: server 1 (10.1.100.22) and server 2 (10.1.100.55).

Alive link health monitor

In this configuration, one server is dead and one server alive. The failed server weight is not over the threshold, so the link health monitor status is alive.

To configure the weight settings on the link health monitor:

1. Configure the link health monitor:

   config system link-monitor
       edit "test-1"
           set srcintf "port22"
           set server-config individual
           set gateway-ip 172.16.202.1
           set failtime 3
           set fail-weight 40
           config server-list
               edit 1
                   set dst "10.1.100.22"
                   set weight 60
               next
               edit 2
                   set dst "10.1.100.55"
                   set weight 30
               next
           end
       next
   end

2. Trigger server 2 to go down. The link monitor is still alive because the fail weight threshold has not been reached.
3. Verify the link health monitor status:

   # diagnose sys link-monitor status test-1
   Link Monitor: test-1, Status: alive, Server num(2), HA state: local(alive), shared(alive)
   Flags=0x1 init, Create time: Fri Jun  4 17:23:29 2021
   Source interface: port22 (14)
   Gateway: 172.16.202.1
   Interval: 500 ms
   Service-detect: disable
   Diffservcode: 000000
   Class-ID: 0
   Fail-weight (40): not activated
     Peer: 10.1.100.22(10.1.100.22)
           Source IP(172.16.202.2)
           Route: 172.16.202.2->10.1.100.22/32, gwy(172.16.202.1)
           protocol: ping, state: alive
                   Latency(Min/Max/Avg): 0.417/0.585/0.530 ms
                   Jitter(Min/Max/Avg): 0.007/0.159/0.057
                   Packet lost: 0.000%
                   Number of out-of-sequence packets: 0
                   Fail Times(0/3)
                   Packet sent: 239, received: 236, Sequence(sent/rcvd/exp): 240/240/241
     Peer: 10.1.100.55(10.1.100.55)
           Source IP(172.16.202.2)
           Route: 172.16.202.2->10.1.100.55/32, gwy(172.16.202.1)
           Fail weight 30 applied
           protocol: ping, state: dead
                   Packet lost: 100.000%
                   Number of out-of-sequence packets: 0
                   Recovery times(0/5) Fail Times(1/3)
                   Packet sent: 239, received: 3, Sequence(sent/rcvd/exp): 240/4/5

Dead link health monitor

In this configuration, one server is dead and one server alive. The failed server weight is over the threshold, so the link health monitor status is dead.

To configure the weight settings on the link health monitor:

1. Configure the link health monitor:

   config system link-monitor
       edit "test-1"
           set srcintf "port22"
           set server-config individual
           set gateway-ip 172.16.202.1
           set failtime 3
           set fail-weight 40
           config server-list
               edit 1
                   set dst "10.1.100.22"
                   set weight 30
               next
               edit 2
                   set dst "10.1.100.55"
                   set weight 50
               next
           end
       next
   end

2. Trigger server 2 to go down. The link monitor is dead because the fail weight threshold has been reached.
3. Verify the link health monitor status:

   # diagnose sys link-monitor status  test-1
   Link Monitor: test-1, Status: dead, Server num(2), HA state: local(dead), shared(dead)
   Flags=0x9 init log_downgateway, Create time: Fri Jun  4 17:23:29 2021
   Source interface: port22 (14)
   Gateway: 172.16.202.1
   Interval: 500 ms
   Service-detect: disable
   Diffservcode: 000000
   Class-ID: 0
   Fail-weight (40): activated
     Peer: 10.1.100.22(10.1.100.22)
           Source IP(172.16.202.2)
           Route: 172.16.202.2->10.1.100.22/32, gwy(172.16.202.1)
           protocol: ping, state: alive
                   Latency(Min/Max/Avg): 0.393/0.610/0.520 ms
                   Jitter(Min/Max/Avg): 0.009/0.200/0.095
                   Packet lost: 0.000%
                   Number of out-of-sequence packets: 0
                   Fail Times(0/3)
                   Packet sent: 680, received: 677, Sequence(sent/rcvd/exp): 681/681/682
     Peer: 10.1.100.55(10.1.100.55)
           Source IP(172.16.202.2)
           Route: 172.16.202.2->10.1.100.55/32, gwy(172.16.202.1)
           Fail weight 50 applied
           protocol: ping, state: dead
                   Packet lost: 100.000%
                   Number of out-of-sequence packets: 0
                   Recovery times(0/5) Fail Times(1/3)
                   Packet sent: 680, received: 3, Sequence(sent/rcvd/exp): 681/4/5