I85PREP
⌘ K
FCP study materials
FCP study resources

FCP: High Availability and System Maintenance

FCP · Work through your exam, one topic at a time.

All guide sections

High Availability and System Maintenance

High Availability and System Maintenance tests FortiGate clustering behavior and the routines that keep a device healthy. Questions often target failover triggers, session pickup, configuration synchronization, and upgrade practice.

HA fundamentals

  • FortiGate Clustering Protocol supports active-passive and active-active; active-passive is common for UTM inspection.
  • Members exchange heartbeats on HA interfaces; monitored interfaces can trigger failover if they go down.
  • Config sync replicates most configuration from primary to secondary; some settings are per-device and do not sync.
  • Session pickup preserves TCP sessions during failover; without it, sessions are dropped and clients retry.
  • Virtual MAC addresses move with the primary role; ARP tables on peers may need to expire.
  • Override lets a higher-priority member preempt; without override, the current primary stays primary.

Maintenance

  • Back up configuration before upgrades; restore validates the config version.
  • Firmware upgrades should follow the upgrade path; HA upgrades can use uninterruptible or interruptible modes.
  • Factory reset clears configuration; keep a backup and know default admin access.
  • FortiGuard updates require connectivity; scheduled updates reduce exposure windows.
  • HA heartbeat can be unicast or multicast; Layer 2 connectivity and switch behavior matter.

Worked example

An active-passive cluster fails over, but users lose SSH sessions while HTTPS reconnects. Session pickup is enabled for TCP, but the SSH sessions were established before the secondary synchronized the session table. On failover, the new primary does not have those sessions, so they reset. HTTPS reconnects automatically, masking the issue. Check HA status and session pickup settings, and verify that sessions existed on both members. Session pickup depends on synchronization timing and protocol state; it does not save every session.

Traps

  • Config sync does not replicate every per-device setting.
  • Active-active does not double throughput for a single flow; it load-balances sessions.
  • Monitored interface failure can cause split-brain if heartbeat and data paths are not planned.

Key points

  • FGCP supports active-passive and active-active clustering.
  • Session pickup preserves sessions only when state is synchronized.
  • Config sync replicates most, but not all, settings.
  • Override controls whether a higher-priority member preempts the primary.
  • HA upgrades and firmware changes require a supported upgrade path.