High availability: two firewalls, no outage
▶ Start the interactive lesson
At 2pm on a Friday the head office firewall's power supply dies. Every site goes dark for 40 minutes while someone drives in with a spare. The board approves a second box. "This time," Maya says, "the spare is already running, already has the config, and already knows every open session."
Key points
- Active-passive: one unit forwards, the other waits with a synced config and session table. Simple and the most common. Active-active: both forward, giving more throughput but more complex (asymmetric flows, more troubleshooting).
- Dedicated HA links carry heartbeats (are you alive?), config sync and session sync. Use two heartbeat links on separate paths. If heartbeats fail but both units are alive, you get split brain, with two actives fighting.
- Failover triggers: the unit dies, a monitored interface goes down, or a monitored path (e.g. ping to the ISP) fails. Priority and preemption decide whether the original primary takes back control when it recovers.
- Related tools: clustering (many firewalls acting as one, for scale), link aggregation to switches, and ECMP / route tracking for path redundancy.
- Test failover in a change window, and upgrade HA pairs one unit at a time.
The full lesson has more: animations, a hands-on exercise, 3 quiz questions and spaced review cards. Open it free →