I have a simple situation – I have two ISP’s. The gateway of my primary (i.e. providing the fastest service) is 9.9.9.9. The gateway of my secondary isp is 8.8.8.8

# My default route via my primary ISP
set routing-options static route 0.0.0.0/0 next-hop 9.9.9.9


set services rpm probe FAILOVER1 test probe-ge1 probe-type icmp-ping
set services rpm probe FAILOVER1 test probe-ge1 target address 8.8.8.8
set services rpm probe FAILOVER1 test probe-ge1 probe-count 5
set services rpm probe FAILOVER1 test probe-ge1 probe-interval 1
set services rpm probe FAILOVER1 test probe-ge1 test-interval 5
set services rpm probe FAILOVER1 test probe-ge1 thresholds total-loss 3
# I have my main service provider on irb.20
set services rpm probe FAILOVER1 test probe-ge1 destination-interface irb.20
set services rpm probe FAILOVER1 test probe-ge1 next-hop 9.9.9.9




set services ip-monitoring policy GE1 match rpm-probe FAILOVER1
# if my main service provider goes down - send traffic through my secondary ISP
set services ip-monitoring policy GE1 then preferred-route route 0.0.0.0/0 next-hop 8.8.8.8
set services ip-monitoring policy GE1 then preferred-route route 0.0.0.0/0 preferred-metric 4