Deployment and System Configuration
Deployment and System Configuration
This domain tests whether you can take a FortiGate from the box to a working, manageable, licensed device - and prove you understand the settings everything else depends on. Expect questions on operating modes, interface types, administrative hardening, registration, firmware upgrade practice and configuration backup.
Core rules
- Operating mode. NAT mode (default) routes between interfaces; transparent mode (
config system settings,set opmode transparent) bridges them, needs a management IP (set manageip), and matches policies on the original ingress interface. - Interface types. Physical, VLAN (
set vlanid), aggregate (LACP), redundant (active/backup), software switch, zone (config system zone), loopback and tunnel. A zone lets one policy cover several interfaces. - Admin access. Each interface carries an
allowaccesslist (ping, https, ssh, fgfm, snmp). Trim it on WAN-facing ports and use trusted hosts so an admin account only authenticates from known subnets. - Registration. Register to FortiCare for FortiGuard updates, support and firmware downloads; a FortiGate VM needs a licence or FortiFlex entitlement before it forwards production traffic.
- Firmware. Follow the documented upgrade path, back up the config and record the current build. Downgrades usually require restoring a backup made on the matching version.
- Backups and revisions.
execute backup config tftp|usb|ftpwrites a full config (plain or encrypted); enable revision backup so earlier configurations are recoverable.
Worked example
A branch FortiGate: port1 to the ISP as a DHCP client with only ping allowed; port2 as the LAN at 10.10.0.1/24 with a DHCP server and DNS; default route learned from DHCP; DNS and NTP pointed at internal servers; admin trusted hosts pinned to 10.10.0.0/24 plus a jump host. Reasoning: the WAN interface keeps only what the ISP needs, management rides the LAN, and time synchronisation must be correct before certificates, logs and FSSO behave.
Traps
Removing https/ssh from every interface before trusted hosts are set, then locking yourself out; forgetting that transparent mode still needs a management IP; restoring a configuration from a different model or firmware branch; upgrading an HA pair without reading the upgrade path.