High Availability with VRRP
Version key: π’ = 1.3+ Β· π‘ = 1.4+ Β· π£ = 1.5+
VRRP (Virtual Router Redundancy Protocol) lets two or more VyOS routers share a virtual IP. If the master fails, a backup takes over β clients donβt notice.
Architecture
βββββββββββββββββββββββββββββββ
β Virtual IP: .1 β
β βββββββββββ βββββββββββ β
β β Router A β β Router B β β
β β MASTER β β BACKUP β β
β β .2 β β .3 β β
β ββββββ¬ββββββ ββββββ¬ββββββ β
β β β β
βββββββββΌββββββββββββββΌββββββββββ
β SWITCH β
βββββββ΄ββββββββββββββ΄ββββββ
β 192.168.1.0/24 β
β Gateway: 192.168.1.1 β β Virtual IP
βββββββββββββββββββββββββββBasic VRRP π’ 1.3+
Router A (master):
set interfaces ethernet eth1 address 192.168.1.2/24
set interfaces ethernet eth1 vrrp vrrp-group 1
set high-availability vrrp group LAN interface eth1
set high-availability vrrp group LAN vrid 10
set high-availability vrrp group LAN virtual-address 192.168.1.1/24
set high-availability vrrp group LAN priority 200 # higher = master
set high-availability vrrp group LAN preempt true
set high-availability vrrp group LAN preempt-delay 60
set high-availability vrrp group LAN hello-interval 1
set high-availability vrrp group LAN authentication type plaintext-password
set high-availability vrrp group LAN authentication plaintext-password 'vrrp-secret'Router B (backup):
set interfaces ethernet eth1 address 192.168.1.3/24
set interfaces ethernet eth1 vrrp vrrp-group 1
set high-availability vrrp group LAN interface eth1
set high-availability vrrp group LAN vrid 10
set high-availability vrrp group LAN virtual-address 192.168.1.1/24
set high-availability vrrp group LAN priority 100 # lower = backup
set high-availability vrrp group LAN authentication type plaintext-password
set high-availability vrrp group LAN authentication plaintext-password 'vrrp-secret'Multiple VRIDs (Load Sharing)
Split traffic across both routers for active-active:
# Router A β master for vrid 10, backup for vrid 20
set high-availability vrrp group LAN-A interface eth1
set high-availability vrrp group LAN-A vrid 10
set high-availability vrrp group LAN-A virtual-address 192.168.1.1/24
set high-availability vrrp group LAN-A priority 200
set high-availability vrrp group LAN-B interface eth1
set high-availability vrrp group LAN-B vrid 20
set high-availability vrrp group LAN-B virtual-address 192.168.1.2/24
set high-availability vrrp group LAN-B priority 100
# Router B β master for vrid 20, backup for vrid 10
# Set priorities reversedLAN clients use .1 or .2 as gateway β split them via DHCP or manual config.
WAN Failover (Dual WAN)
If each router has its own WAN:
# On both routers β SNAT to their respective WAN IPs
# Use `translation address` instead of masquerade
# Router A
set nat source rule 100 outbound-interface name eth0
set nat source rule 100 source address 192.168.1.0/24
set nat source rule 100 translation address 203.0.113.10
# Router B
set nat source rule 100 outbound-interface name eth0
set nat source rule 100 source address 192.168.1.0/24
set nat source rule 100 translation address 203.0.113.11Configuration Sync
VRRP handles IP failover, but you need config sync too. Options:
Manual Sync (simple)
# On master, after commit:
save
scp /config/config.boot backup@192.168.1.3:/config/config.boot
ssh backup@192.168.1.3 'load /config/config.boot; commit; save'Scripted Sync
Create /config/scripts/sync-config.sh:
#!/bin/bash
BACKUP="192.168.1.3"
REMOTE_USER="vyos"
CONFIG="/config/config.boot"
rsync -avz $CONFIG $REMOTE_USER@$BACKUP:/config/config.boot.tmp
ssh $REMOTE_USER@$BACKUP "source /opt/vyatta/etc/functions/script-template; \
configure; load /config/config.boot.tmp; \
set high-availability vrrp group LAN priority 100; \
commit; save; exit"
Cluster Config (VyOS 1.4+) π‘
# Experimental: config-sync feature
set high-availability config-sync peer 192.168.1.3
set high-availability config-sync sync-direction master-to-backup
set high-availability config-sync sync-on-commitStateful Failover (Connection Tracking Sync) π‘ 1.4+
For seamless failover of active connections (NAT, firewall states):
# On master β sync state table to backup
set high-availability vrrp group LAN sync-group LAN-SYNC
set high-availability vrrp sync-group LAN-SYNC member LAN
set high-availability vrrp sync-group LAN-SYNC conntrack-sync interface eth1
# On backup β receive
set high-availability vrrp group LAN sync-group LAN-SYNC
set high-availability vrrp sync-group LAN-SYNC member LAN
set high-availability vrrp sync-group LAN-SYNC conntrack-sync interface eth1
# Note: Both routers need a dedicated sync link (or use LAN).
# The sync interface carries state table traffic.Monitoring & Health Checks π‘ 1.4+
Track Interface Status
Drop priority if WAN goes down:
set high-availability vrrp group LAN track interface eth0
set high-availability vrrp group LAN track interface eth0 priority-cost 150
# If eth0 goes down, priority drops by 150 β backup takes overScript-based Health Check
# Health check script
set high-availability vrrp group LAN health-check script /config/scripts/health-check.sh
set high-availability vrrp group LAN health-check interval 5
# /config/scripts/health-check.sh:#!/bin/bash
# Exit 0 = healthy, exit 1 = unhealthy (trigger failover)
ping -c 1 -W 2 8.8.8.8 > /dev/null 2>&1
exit $?Verify
# VRRP status
show high-availability vrrp summary
show high-availability vrrp group LAN
show high-availability vrrp statistics
# Who is master?
show high-availability vrrp group LAN | grep -i state
# Conntrack sync status
show high-availability vrrp sync-group
# Interface tracking
show high-availability vrrp trackComplete HA Config (Reference)
Two routers, shared LAN, dual WAN, config sync:
Router A (192.168.1.2) Router B (192.168.1.3)
WAN: 203.0.113.10 WAN: 203.0.113.11
Priority: 200 Priority: 100
Virtual IP: 192.168.1.1 Virtual IP: 192.168.1.1Both routers share the same base config except:
- Real IP on eth1 (.2 vs .3)
- VRRP priority (200 vs 100)
- SNAT WAN IP
Template approach: maintain ONE config, deploy with sed to customize per-router values.
Pitfalls
- Split brain: Both routers master β both respond to virtual IP ARP. Caused by: no multicast between them, mismatched VRID, authentication mismatch.
- No state sync = broken connections during failover. NAT tables arenβt shared by default β enable conntrack-sync.
- STP convergence on switch ports can delay failover. Use PortFast/edge ports.
- WAN failover β link redundancy unless you have BGP with your own IP space. Two different WAN IPs mean two different NAT IPs.