Demo – made-up data. Northline, the sites and the numbers are invented. Notices 3.5 – work in progress
VANTAGENorthline
Collected✕7 sites need attention

Action list

Guide: Service desk › 5. When to pass a call onEverything to address, at every site, and how to fix it · built Sun Nov 22 2026 04:16 PM
  • ✕Philadelphiaunreachable
    • Collection failing since Nov 22, 11:16 AM (4 attempts): timed out – showing data from Nov 20, 2:16 PM
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    collector
    timed out  ·  TCP 22: no answer on port 22
    How to fix
    No other site has a tunnel up to Philadelphia (2 configured) – the site may really be down.

    The collector couldn't log in to this FortiGate this time (timed out), so the page shows the last good data. It has failed 4 times in a row since Nov 22, 11:16 AM.

    1. Check the site is up some other way.
    2. From NET-DEMO01: test the path (below).
    3. Read the FortiGate's event log for the login.
    # on NET-DEMO01 (PowerShell):
    Test-NetConnection 10.250.0.18 -Port 22
    
    # on the FortiGate (your admin account) – why it refused the login:
    execute log filter category event
    execute log filter field user netmap
    execute log display
    
    # --- on Philadelphia's FortiGate. Safe to share – no keys:
    get router info routing-table details 10.250.0.5
    diagnose sniffer packet any 'host 10.250.0.5 and port 22' 4 20
    #   nothing arrives          -> the request never left the far end on a tunnel that reaches this site
    #   arrives, no reply leaves -> this FortiGate ignores it: policy, or the source is no longer the trusted host
    #   reply leaves elsewhere   -> the return route is wrong here
    
    # The next commands print LIVE KEYS (tunnel encryption keys, pre-shared keys). Read them on the FortiGate –
    # never paste their output into a ticket, an email or a chat.
    #   diagnose vpn tunnel list   ·   diagnose vpn ike gateway list   ·   show vpn ipsec phase1-interface
  • ▲Houston OTPLC-LINE2 unreachable
    • Collector ping to PLC-LINE2 (10.3.90.5) failed: no reply
    Explain →
    How to fix

    The collector cannot ping PLC-LINE2 (10.3.90.5). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲PhiladelphiaFG-PHILADELPH-SW1 unreachable
    • Collector ping to FG-PHILADELPH-SW1 (10.255.18.10) failed: no reply
    Explain →
    How to fix

    The collector cannot ping FG-PHILADELPH-SW1 (10.255.18.10). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲PhiladelphiaPhiladelphia FortiGate unreachable
    • Collector ping to Philadelphia FortiGate (10.250.0.18) failed: no reply
    Explain →
    How to fix

    The collector cannot ping Philadelphia FortiGate (10.250.0.18). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Philadelphiaphiladelphia FortiGate (users) unreachable
    • Collector ping to philadelphia FortiGate (users) (10.18.10.1) failed: no reply
    Explain →
    How to fix

    The collector cannot ping philadelphia FortiGate (users) (10.18.10.1). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Philadelphiaphiladelphia servers gateway unreachable
    • Collector ping to philadelphia servers gateway (10.18.30.1) failed: no reply
    Explain →
    How to fix

    The collector cannot ping philadelphia servers gateway (10.18.30.1). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Philadelphiaphiladelphia printers gateway unreachable
    • Collector ping to philadelphia printers gateway (10.18.40.1) failed: no reply
    Explain →
    How to fix

    The collector cannot ping philadelphia printers gateway (10.18.40.1). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Philadelphiaphiladelphia cameras gateway unreachable
    • Collector ping to philadelphia cameras gateway (10.18.50.1) failed: no reply
    Explain →
    How to fix

    The collector cannot ping philadelphia cameras gateway (10.18.50.1). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Philadelphiaphiladelphia iot gateway unreachable
    • Collector ping to philadelphia iot gateway (10.18.90.1) failed: no reply
    Explain →
    How to fix

    The collector cannot ping philadelphia iot gateway (10.18.90.1). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Philadelphiaphiladelphia FortiLink unreachable
    • Collector ping to philadelphia FortiLink (10.255.18.1) failed: no reply
    Explain →
    How to fix

    The collector cannot ping philadelphia FortiLink (10.255.18.1). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲PhiladelphiaFG-PHILADELPH-AP1 unreachable
    • Collector ping to FG-PHILADELPH-AP1 (10.18.10.20) failed: no reply
    Explain →
    How to fix

    The collector cannot ping FG-PHILADELPH-AP1 (10.18.10.20). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲PhiladelphiaFG-PHILADELPH-AP2 unreachable
    • Collector ping to FG-PHILADELPH-AP2 (10.18.10.21) failed: no reply
    Explain →
    How to fix

    The collector cannot ping FG-PHILADELPH-AP2 (10.18.10.21). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Philadelphiaphiladelphia DNS 10.1.30.2 unreachable
    • Collector ping to philadelphia DNS 10.1.30.2 (10.1.30.2) failed: no reply
    Explain →
    How to fix

    The collector cannot ping philadelphia DNS 10.1.30.2 (10.1.30.2). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Philadelphiaphiladelphia DNS 10.1.30.3 unreachable
    • Collector ping to philadelphia DNS 10.1.30.3 (10.1.30.3) failed: no reply
    Explain →
    How to fix

    The collector cannot ping philadelphia DNS 10.1.30.3 (10.1.30.3). A critical internal address is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ●Atlanta2 APs disconnected
    • 2 APs disconnected: FG-ATLANTA-AP15, FG-ATLANTA-AP16
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get wireless-controller wtp-status
    WTP: FP231F09150001
        wtp-id           : FP231F09150001
        name             : FG-ATLANTA-AP15
        connection-state : Disconnected
        join-time        : Thu Nov 12 08:00:00 2026
    
    WTP: FP231F09160001
        wtp-id           : FP231F09160001
        name             : FG-ATLANTA-AP16
        connection-state : Disconnected
        join-time        : Thu Nov 12 08:00:00 2026
    How to fix

    The access point isn't talking to the FortiGate's wireless controller, so it isn't serving Wi-Fi. Usually power (PoE), cabling, or the AP itself.

    1. Check PoE on its switch port (FG-ATLANTA-SW5 port5) – if the port shows no power or 100 Mb, check the cable.
    2. Power-cycle the port (restarts the AP).
    3. Still down: check it on site (lights), or swap the AP.
    # The switch and port below are what the access point ITSELF announced (LLDP) - they were not read from the switch.
    # Confirm them before resetting power: the Find page, or the first command.
    diagnose switch-controller switch-info lldp neighbors-summary
    diagnose switch-controller switch-info poe summary FG-ATLANTA-SW5
    diagnose switch-controller switch-info port-stats FG-ATLANTA-SW5 port5
    execute switch-controller poe-reset FG-ATLANTA-SW5 port5
    get wireless-controller wtp-status
  • ●AtlantaDHCP scope filling up
    • DHCP filling up: users – 18 of 20 addresses in use (10% free) (estimate: seen at the last collection, plus reservations)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system dhcp server
    config system dhcp server
        edit 1
            set default-gateway 10.9.10.1
            set netmask 255.255.255.0
            set interface "users"
            set dns-server 10.1.30.2 10.1.30.3
            set ntp-server 10.1.30.2 10.1.30.3
            set domain-name "northline.example.com"
            set lease-time 28800
            end
        next
    end
    How to fix

    Almost every address in this DHCP scope is taken. New or returning devices on that network get no address – they look as if the network is down. The count is an estimate from what was seen at the last collection plus the reservations; leases held by devices that are off make the real picture worse, not better.

    1. See who holds the leases (below, with an admin account) – look for stale entries and devices that do not belong there.
    2. Shorten the lease time on networks where devices come and go (guest, Wi-Fi).
    3. Widen the range if the subnet has room, or move devices to another network.
    4. Clear leases that are clearly stale.
    execute dhcp lease-list users
    
    # shorter lease / wider range:
    show system dhcp server
    
    # clear one stale lease:
    execute dhcp lease-clear <ip>

    Networks & VLANs →

  • ●Dallas1 switch offline
    • 1 switch offline: FG-DALLAS-SW6
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    diagnose switch-controller switch-info status
    Not in the output (the FortiGate got no answer from): FG-DALLAS-SW6
    How to fix

    The FortiGate can't reach this FortiSwitch over FortiLink. Everything plugged into it – and any switch behind it – is probably offline as well.

    1. Check the switch has power and its uplink link light is on (on site).
    2. From the FortiGate, check the switch and the LLDP neighbours of the switch above it.
    3. If it was removed or replaced on purpose, delete it from the FortiGate or acknowledge it in the site file.
    diagnose switch-controller switch-info status FG-DALLAS-SW6
    execute switch-controller get-conn-status
    diagnose switch-controller switch-info lldp neighbors-summary
    
    # removed for good:
    config switch-controller managed-switch
        delete "FG-DALLAS-SW6"
    end
  • ●DallasDHCP scope nearly full
    • DHCP nearly full: cameras – 10 of 10 addresses in use (0% free) (estimate: seen at the last collection, plus reservations)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system dhcp server
    config system dhcp server
        edit 5
            set default-gateway 10.8.50.1
            set netmask 255.255.255.0
            set interface "cameras"
            set dns-server 10.1.30.2 10.1.30.3
            set ntp-server 10.1.30.2 10.1.30.3
            set domain-name "northline.example.com"
            end
        next
    end
    How to fix

    Almost every address in this DHCP scope is taken. New or returning devices on that network get no address – they look as if the network is down. The count is an estimate from what was seen at the last collection plus the reservations; leases held by devices that are off make the real picture worse, not better.

    1. See who holds the leases (below, with an admin account) – look for stale entries and devices that do not belong there.
    2. Shorten the lease time on networks where devices come and go (guest, Wi-Fi).
    3. Widen the range if the subnet has room, or move devices to another network.
    4. Clear leases that are clearly stale.
    execute dhcp lease-list cameras
    
    # shorter lease / wider range:
    show system dhcp server
    
    # clear one stale lease:
    execute dhcp lease-clear <ip>

    Networks & VLANs →

  • ●Chicago1 port forward to review

    Published to the internet. Confirm each is still needed and limited to who needs it.

    • web (203.0.113.11 tcp 8443 → 10.1.30.20)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show firewall vip
    config firewall vip
        edit "web"
            set extip 203.0.113.11
            set mappedip "10.1.30.20"
            set portforward enable
            set extport 8443
        next
    end
    How to fix

    These addresses/ports on a public IP are forwarded to internal devices. Each is only reachable if a firewall policy allows it – review that each is still needed and limited to the sources that need it.

    1. Check which firewall policy uses each one (below) and who its source is.
    2. Not needed: delete the policy, then the VIP.
    3. Needed: limit the source (set src-filter, or a source address group in the policy) – or use the VPN instead.
    4. Cameras / DVRs: use the vendor's cloud relay or the VPN rather than an open port.

    web: 203.0.113.11 tcp 8443 → 10.1.30.20

    show firewall policy | grep -f "web"
    
    config firewall vip
        edit "web"
            set src-filter <allowed public IPs>
        next
    end
  • ●Chicagoknown-bad address
    • 198.51.100.96/28 – Address object: address Partner-FTP-Net (used in policy 20)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show firewall address
    config firewall address
        edit "Partner-FTP-Net"
            set uuid 89683144-2c32-ef70-ce6f-acad015e2c6c
            set subnet 198.51.100.96 255.255.255.240
            set comment "Old EDI partner"
        next
    end
    How to fix

    An address this FortiGate is configured to trust or talk to is on a public list of compromised or hostile hosts (Emerging Threats, rule numbers 2000000-2799999). A list is not proof: addresses are re-used, and a listed address may belong to someone else today.

    1. Find out who the address belongs to today (whois) and whether anyone still needs it.
    2. Look at the traffic to and from it in the FortiGate's logs (below; read-only).
    3. Not needed: remove it from the object, tunnel or neighbour. Needed and checked: acknowledge it in the site file ("acknowledged": {"badip_match": {"198.51.100.96/28": "why"}}).

    198.51.100.96/28: address Partner-FTP-Net

    execute log filter reset
    execute log filter category traffic
    execute log filter field srcip 198.51.100.96/28
    execute log display
    
    execute log filter reset
    execute log filter category traffic
    execute log filter field dstip 198.51.100.96/28
    execute log display
  • ●Dallasadmin HTTPS allows old TLS
    • admin-https-ssl-versions still includes tlsv1-1 (today: tlsv1-1 tlsv1-2). Prefer tlsv1-2 tlsv1-3.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system global
    admin-https-redirect    : enable
    admin-https-ssl-versions: tlsv1-1 tlsv1-2
    admin-lockout-duration  : 60
    admin-lockout-threshold : 3
    admin-server-cert       : Fortinet_Factory
    admin-ssh-v1            : disable
    admin-telnet            : disable
    strong-crypto           : enable
    How to fix

    The HTTPS admin GUI still offers old TLS versions (SSLv3 / TLS 1.0 / TLS 1.1). Those are no longer considered safe; keep TLS 1.2 and, on modern FortiOS, TLS 1.3.

    1. Confirm management browsers support TLS 1.2 (and TLS 1.3 if you enable it).
    2. Set admin-https-ssl-versions to tlsv1-2 tlsv1-3 (below).
    3. Open the HTTPS GUI once to confirm the login page still loads.
    get system global | grep admin-https-ssl-versions
    
    config system global
        set admin-https-ssl-versions tlsv1-2 tlsv1-3
    end
  • ●Dallasadmin GUI uses factory certificate
    • admin-server-cert is Fortinet_Factory – browsers warn, and the factory cert is shared across FortiGates. Use a certificate from your CA for the management name.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system global
    admin-https-redirect    : enable
    admin-https-ssl-versions: tlsv1-1 tlsv1-2
    admin-lockout-duration  : 60
    admin-lockout-threshold : 3
    admin-server-cert       : Fortinet_Factory
    admin-ssh-v1            : disable
    admin-telnet            : disable
    strong-crypto           : enable
    How to fix

    The HTTPS admin GUI uses the factory certificate (Fortinet_Factory). Browsers warn, and that default cert is shared across FortiGates – replace it with one from your CA.

    1. Issue a certificate from your CA for the FortiGate's management name (or IP).
    2. Import it on the FortiGate (System → Certificates, or the CLI).
    3. Set admin-server-cert to that certificate (below).
    4. Confirm the browser shows your CA as the issuer and no warning.
    get vpn certificate local details
    get system global | grep admin-server-cert
    
    config system global
        set admin-server-cert "<your-ca-cert>"
    end
  • ●Dallasadmin lockout is short
    • Sign-in lockout duration is under 5 minutes (threshold 3 · duration 60). A short lockout barely slows password guessing.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system global
    admin-https-redirect    : enable
    admin-https-ssl-versions: tlsv1-1 tlsv1-2
    admin-lockout-duration  : 60
    admin-lockout-threshold : 3
    admin-server-cert       : Fortinet_Factory
    admin-ssh-v1            : disable
    admin-telnet            : disable
    strong-crypto           : enable
    How to fix

    Admin sign-in lockout is on, but the lockout duration is under five minutes. That barely slows someone guessing passwords. Fortinet's hardening examples use around 15 minutes.

    1. Pick a duration of at least 300 seconds (900 is a common choice).
    2. Set admin-lockout-duration (below) – keep a non-zero threshold.
    3. Confirm a failed login still locks for the new period.
    get system global | grep admin-lockout
    
    config system global
        set admin-lockout-duration 900
    end
  • ●Dallasknown-exploited vulnerability
    • FortiGate firmware 7.2.4 on FG-DALLAS is listed as affected by 1 vulnerability that is being exploited in the wild (used by attackers elsewhere – not a sign that this device was attacked): CVE-2024-55501 – matched by version; check the advisory
    Explain →
    How to fix

    This firmware version is on the vendor's affected list for a vulnerability that CISA records as exploited in real attacks. The match is by version only: a vulnerability can depend on a feature you do not use (SSL-VPN, the admin interface on the internet, a particular model) – the advisory says which.

    1. Read the advisory for each one (links on the Firmware page) – does it apply to how this device is set up?
    2. Plan the upgrade to a version that is no longer listed ('Fixed in' on the Firmware page).
    3. If one does not apply, acknowledge it with the reason so it stops showing: for every site in config.json, or for this site only in its site file (both below). It stays on the Firmware page, marked.

    CVE-2024-55501 · CVSS 9.8 · fixed in 7.2.5 – Made-up advisory used so the demo can show a vulnerability marked exploited in the wild.

    get system status
    
    # to acknowledge for EVERY site, in config.json:
    "security": {"vuln_acknowledged": {"CVE-0000-00000": "why it does not apply"}}
    
    # or for this site only, in sites\dallas.json:
    "acknowledged": {"vuln_kev": {"CVE-0000-00000": "why it does not apply here"}}

    Known vulnerabilities →

  • ●MinneapolisWi-Fi security
    • Northline
      • WPA2-Personal only – a shared password; WPA3-SAE (or transition) is stronger
      • PMF off
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show wireless-controller vap (no 'set security' line = the default, WPA2-Personal)
    config wireless-controller vap
        edit "corp"
            set ssid "Northline"
            set security wpa2-only-personal
            set pmf disable
            set local-bridging enable
        next
    end
    How to fix

    Wi-Fi settings below Fortinet's recommendations. WPA2-Personal shares one password with everyone (and anyone who has ever had it); without PMF, clients can be knocked off with forged de-auth frames; guests who can reach each other can attack each other's devices.

    1. Corporate SSIDs: WPA2/WPA3-Enterprise with RADIUS, PMF optional or required.
    2. Shared-password SSIDs: WPA3-SAE transition (wpa3-sae-transition), PMF optional – check old devices first.
    3. Guest: enable intra-vap-privacy (client isolation).
    4. Test with one SSID before changing all sites.

    Northline: WPA2-Personal only – a shared password; WPA3-SAE (or transition) is stronger, PMF off

    config wireless-controller vap
        edit "corp"
            set pmf optional
        next
    end
  • ●New Yorkknown-bad addresses not blocked
    • No firewall policy drops FortiGuard's known-bad addresses (Malicious servers, Botnet C&C, Tor exit nodes) for traffic coming in from the internet or traffic going out to them. SD-WAN is on: a policy names the zone (virtual-wan-link), not its member ports.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show firewall policy
    (no policy uses FortiGuard Internet Service entries)
    How to fix

    FortiGuard already knows the addresses of malicious servers, botnet controllers and Tor exit nodes (the Internet Service Database, part of the FortiGuard subscription). A deny policy that uses them stops known-bad traffic at the edge – the FortiGate does the blocking; Vantage only reads the configuration. The policy only works above the allow policies: FortiGate uses the first policy that matches. SD-WAN is on, so the policies name the SD-WAN zone (virtual-wan-link: wan1, wan2). A member port cannot be used in a policy on its own.

    1. Check the Internet Service names on this firmware (first line below) and adjust them if they differ.
    2. Block-known-bad-in is added as policy 12, the next free number. No allow policy uses the internet side yet, so it does not need moving.
    3. Block-known-bad-out is added as policy 13, the next free number. The move line puts it above allow policy 1, the first one that lets internet traffic through.
    4. Paste the commands in a change window. The last lines show each policy, so you can see it took.
    5. Watch the logs for a day: if something you need is blocked, find out why before making an exception.
    6. Blocking another way (an external threat feed, or upstream)? Acknowledge this note in the site file ("acknowledged": {"badip_block": {"fortiguard-block": "why"}}).

    Caution: a deny policy placed above your allow policies can stop traffic you need if FortiGuard lists an address you use. Use a change window and keep console or out-of-band access.

    # The names differ between firmware versions - check them first:
    show firewall internet-service-name | grep -i -e malicious -e botnet -e tor
    
    config firewall policy
        edit 12
            set name "Block-known-bad-in"
            set srcintf "virtual-wan-link"
            set dstintf "any"
            set internet-service-src enable
            set internet-service-src-name "Malicious-Malicious.Server" "Botnet-C&C.Server" "Tor-Exit.Node"
            set dstaddr "all"
            set action deny
            set schedule "always"
            set service "ALL"
            set logtraffic all
        next
        edit 13
            set name "Block-known-bad-out"
            set srcintf "any"
            set dstintf "virtual-wan-link"
            set srcaddr "all"
            set internet-service enable
            set internet-service-name "Malicious-Malicious.Server" "Botnet-C&C.Server"
            set action deny
            set schedule "always"
            set logtraffic all
        next
        move 13 before 1
    end
    
    # check the result:
    show firewall policy 12
    show firewall policy 13
  • ●Phoenixadmin open to internet
    • Admin login reachable on a public address: wan1 203.0.113.27 (static) allows https/ssh
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system interface
    config system interface
        edit "wan1"
            set ip 203.0.113.27 255.255.255.248
            set allowaccess ping https ssh
            set role wan
        next
    end
    How to fix

    This interface has a public IP address and allows admin access, so the FortiGate's login is reachable from the internet (the Evidence shows the address and the allowaccess line). Trusted hosts limit who can log in, but the login page is still exposed to scanning and to any future vulnerability in it. Manage over the VPN / internal interfaces instead.

    1. Confirm nobody manages this FortiGate through these interfaces (check with the team first).
    2. Set allowaccess to the remaining services only (below keeps everything else as it is).
    config system interface
        edit "wan1"
            set allowaccess ping
        next
    end
  • ●Phoenixrisky service exposed
    • 1 port forward(s) put RDP on a public address: rdp-jump (203.0.113.27 tcp 3389 → 10.17.30.40) – reachable from the internet if a firewall policy allows it
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show firewall vip
    config firewall vip
        edit "rdp-jump"
            set extip 203.0.113.27
            set mappedip "10.17.30.40"
            set extintf "wan1"
            set portforward enable
            set extport 3389
            set mappedport 3389
        next
    end
    How to fix

    These port forwards publish an internal server's management or file-sharing service on a public address. RDP, SSH, SMB and database ports are the most scanned and brute-forced services on the internet.

    1. Check which firewall policy uses each one (below) and who its source is.
    2. Not needed: delete the policy, then the VIP.
    3. Needed: limit the source (set src-filter, or a source address group in the policy) – or use the VPN instead.
    4. Cameras / DVRs: use the vendor's cloud relay or the VPN rather than an open port.

    rdp-jump: 203.0.113.27 tcp 3389 → 10.17.30.40

    show firewall policy | grep -f "rdp-jump"
    
    config firewall vip
        edit "rdp-jump"
            set src-filter <allowed public IPs>
        next
    end
  • ●Phoenixknown-bad addresses not blocked
    • No firewall policy drops FortiGuard's known-bad addresses (Malicious servers, Botnet C&C) for traffic going out to them. SD-WAN is on: a policy names the zone (virtual-wan-link), not its member ports.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show firewall policy
            set internet-service-src enable
            set internet-service-src-name "Malicious-Malicious.Server" "Botnet-C&C.Server" "Tor-Exit.Node"
    How to fix

    FortiGuard already knows the addresses of malicious servers, botnet controllers and Tor exit nodes (the Internet Service Database, part of the FortiGuard subscription). A deny policy that uses them stops known-bad traffic at the edge – the FortiGate does the blocking; Vantage only reads the configuration. The policy only works above the allow policies: FortiGate uses the first policy that matches. SD-WAN is on, so the policies name the SD-WAN zone (virtual-wan-link: wan1, wan2). A member port cannot be used in a policy on its own.

    1. Check the Internet Service names on this firmware (first line below) and adjust them if they differ.
    2. Block-known-bad-out is added as policy 13, the next free number. The move line puts it above allow policy 2, the first one that lets internet traffic through.
    3. Paste the commands in a change window. The last lines show each policy, so you can see it took.
    4. Watch the logs for a day: if something you need is blocked, find out why before making an exception.
    5. Blocking another way (an external threat feed, or upstream)? Acknowledge this note in the site file ("acknowledged": {"badip_block": {"fortiguard-block": "why"}}).

    Caution: a deny policy placed above your allow policies can stop traffic you need if FortiGuard lists an address you use. Use a change window and keep console or out-of-band access.

    # The names differ between firmware versions - check them first:
    show firewall internet-service-name | grep -i -e malicious -e botnet -e tor
    
    config firewall policy
        edit 13
            set name "Block-known-bad-out"
            set srcintf "any"
            set dstintf "virtual-wan-link"
            set srcaddr "all"
            set internet-service enable
            set internet-service-name "Malicious-Malicious.Server" "Botnet-C&C.Server"
            set action deny
            set schedule "always"
            set logtraffic all
        next
        move 13 before 2
    end
    
    # check the result:
    show firewall policy 13
  • ▲Bostonmost tunnels down
    • 2 of 3 IPsec tunnels are down: to-chicago, to-chicago-b
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn ipsec tunnel summary
    'to-chicago' 203.0.113.11:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    'to-chicago-b' 198.51.100.11:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    How to fix

    Some IPsec tunnels to other sites are down. If the aggregate's other member is up, traffic still flows – this is lost redundancy, not an outage. Backup tunnels whose phase 2 isn't set to auto-negotiate stay down until traffic needs them, so they look down here.

    1. Check the tunnel from this end (and the other site's end) – the WAN page's Tunnel paths tab shows both.
    2. For backup tunnels, enable auto-negotiate on their phase 2 (below) so they stay up.
    3. Tunnels that point at the wrong IP or a site that no longer exists: fix the remote gateway, or delete them.

    to-chicago → 203.0.113.11
    to-chicago-b → 198.51.100.11

    get vpn ipsec tunnel summary
    diagnose vpn tunnel list name to-chicago
    diagnose vpn tunnel list name to-chicago-b

    Tunnel paths →

  • ▲PhiladelphiaExample Fiber gateway unreachable
    • Collector ping to Example Fiber gateway (203.0.113.1) failed: no reply
    Explain →
    How to fix

    The collector cannot ping Example Fiber gateway (203.0.113.1). An ISP gateway or external service is unreachable from the Vantage server.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲PhoenixMetro Cable down
    • WAN Metro Cable (wan2) is down
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    diagnose sys sdwan health-check
    Health Check(Internet):
      Seq(2 wan2): state(dead), packet-loss(0.000%) latency(9.000), jitter(1.000), mos(4.300), bandwidth-up(400000), bandwidth-dw(400000), bandwidth-bi(800000) sla_map=0x0
    How to fix

    The Metro Cable connection is down. Traffic has failed over to the other connection(s), so users are probably fine, but there is no backup right now.

    1. Check the ISP modem/ONT has power and normal lights (on-site contact – see Circuits & contacts).
    2. Check the FortiGate port is up and has an address.
    3. If the line is dead, call the ISP – circuit, account and support numbers are on the Circuits & contacts tab.
    get system interface physical wan2
    diagnose sys sdwan health-check

    Circuits & contacts →

  • ●Atlantatraffic one way only
    • 1 tunnel(s) carried traffic in one direction only in the last 69 minutes: to-chicago (received 2,462 / sent 0)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn ipsec tunnel summary
    'to-chicago' 203.0.113.11:0  selectors(total,up): 1/1  rx(pkt,err): 90000/0  tx(pkt,err): 0/0
    diagnose vpn tunnel list
    name=to-chicago ver=2 serial=1 203.0.113.19:0->203.0.113.11:0 nexthop=203.0.113.17 tun_id=203.0.113.11 tun_id6=::203.0.113.11 dst_mtu=1500 dpd-link=on weight=1
    bound_if=5 lgwy=static/1 tun=intf mode=auto/1 encap=none/552 options[0228]=npu frag-rfc  run_state=0 role=primary accept_traffic=1 overlay_id=0
    
    proxyid_num=1 child_num=0 refcnt=5 ilast=0 olast=1 ad=/0
    stat: rxp=90000 txp=0 rxb=12780000 txb=0
    dpd: mode=on-demand on=1 status=ok idle=20000ms retry=3 count=0 seqno=223
    natt: mode=none draft=0 interval=0 remote_port=0
    fec: egress=0 ingress=0
    proxyid=to-chicago proto=0 sa=1 ref=2 serial=1 auto-negotiate
      src: 0:0.0.0.0-255.255.255.255:0
      dst: 0:0.0.0.0-255.255.255.255:0
      SA:  ref=3 options=18227 type=00 soft=0 mtu=1438 expire=9387/0B replaywin=2048
           seqno=1 esn=0 replaywin_lastseq=00015f90 qat=0 rekey=0 hash_search_len=1
      life: type=01 bytes=0/0 timeout=9387/43200
      dec: spi=24ac0205 esp=aes key=32 <removed>
           ah=sha256 key=32 <removed>
      enc: spi=c242fdcb esp=aes key=32 <removed>
           ah=sha256 key=32 <removed>
      dec:pkts/bytes=90000/12780000, enc:pkts/bytes=0/0
      npu_flag=03 npu_rgwy=203.0.113.11 npu_lgwy=203.0.113.19 npu_selid=0 dec_npuid=1 enc_npuid=1
    ------------------------------------------------------
    How to fix

    These tunnels are up, but traffic only flows one way: this end sends and never receives (or the reverse). That nearly always means the two ends' IPsec aggregates disagree – different mode, different member order, or a member that exists on one end only – so each end picks a different tunnel.

    1. Compare the aggregate at both ends (the matching 'aggregate' item on this page, if there is one, says which).
    2. Make both ends use algorithm redundant with members in the same path order.
    3. Check again: rx and tx should both climb on the same tunnel.
    4. Also check Best Practices for “VPN allow one way only” and “asymmetric routing with one-way traffic”.

    to-chicago: since the last collection received 2,462 / sent 0

    # safe to share (no keys):
    show system ipsec-aggregate
    diagnose sys ipsec-aggregate list
    get vpn ipsec tunnel summary
    
    # The next commands print LIVE KEYS (tunnel encryption keys, pre-shared keys). Read them on the FortiGate –
    # never paste their output into a ticket, an email or a chat.
    diagnose vpn tunnel list name to-chicago

    Tunnel paths →

  • ●Atlantano OSPF neighbour
    • OSPF is set up on to-chicago but no neighbour is up there – routes to the far end come another way, or not at all
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get router info ospf neighbor
    OSPF process 0, VRF 0:
    Neighbor ID     Pri   State           Dead Time   Address         Interface
    10.1.0.1          1   Init/ -         00:00:35    10.254.9.1      to-chicago
    10.1.0.1          1   Full/ -         00:00:31    10.254.9.5      to-chicago-b
    How to fix

    OSPF is configured on this tunnel or aggregate and the tunnel is up, but no neighbour has formed. Usual causes: the aggregate has no running members (see that item), hellos leave on one member and return on another, or the two ends disagree on area, network type, timers or MTU.

    1. If this page also lists an aggregate that is not running, fix that first – OSPF cannot form over it.
    2. Compare the OSPF interface settings at both ends.
    3. Watch whether hellos arrive and leave (below).
    # on the PRIMARY unit (get system status | grep -i 'HA mode')
    # safe to share (no keys):
    get router info ospf neighbor
    get router info ospf interface
    diagnose sniffer packet any 'proto 89' 4 10
  • ●Bostonno tunnel to Chicago
    • No working IPsec tunnel to Chicago – every tunnel to that site is down
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn ipsec tunnel summary
    'to-chicago' 203.0.113.11:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    'to-chicago-b' 198.51.100.11:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    How to fix

    Some IPsec tunnels to other sites are down. If the aggregate's other member is up, traffic still flows – this is lost redundancy, not an outage. Backup tunnels whose phase 2 isn't set to auto-negotiate stay down until traffic needs them, so they look down here.

    1. Check the tunnel from this end (and the other site's end) – the WAN page's Tunnel paths tab shows both.
    2. For backup tunnels, enable auto-negotiate on their phase 2 (below) so they stay up.
    3. Tunnels that point at the wrong IP or a site that no longer exists: fix the remote gateway, or delete them.

    to-chicago → 203.0.113.11
    to-chicago-b → 198.51.100.11

    get vpn ipsec tunnel summary
    diagnose vpn tunnel list name to-chicago
    diagnose vpn tunnel list name to-chicago-b

    Tunnel paths →

  • ●Bostonpreferred tunnel down
    • IPsec aggregate chicago: its first member to-chicago is down (dialling 203.0.113.11) – traffic is on to-dallas
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system ipsec-aggregate
    config system ipsec-aggregate
        edit "chicago"
            set member "to-chicago" "to-dallas"
            set algorithm redundant
        next
    end
    get vpn ipsec tunnel summary
    'to-chicago' 203.0.113.11:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    'to-dallas' 203.0.113.18:0  selectors(total,up): 1/1  rx(pkt,err): 66620/0  tx(pkt,err): 69134/0
    show vpn ipsec phase1-interface
    config vpn ipsec phase1-interface
        edit "to-chicago"
            set interface "wan1"
            set remote-gw 203.0.113.11
            set ike-version 2
        next
    end
    How to fix

    'redundant' sends everything over the first member that is up. The first member here is down, so this site is on its backup path to the far end – and if that path fails too, there is nothing left.

    1. Why is it down? Check the address it dials is still the far end's (an 'unknown address' item on this page means it is not).
    2. If that circuit is gone for good, take the tunnel out of the aggregate at BOTH ends, so the working tunnel is first.
    3. After changing an aggregate: renegotiate its tunnels and check the running aggregate.
    # on the PRIMARY unit (get system status | grep -i 'HA mode')
    # safe to share (no keys):
    get vpn ipsec tunnel summary
    diagnose sys ipsec-aggregate list
    
    # The next commands print LIVE KEYS (tunnel encryption keys, pre-shared keys). Read them on the FortiGate –
    # never paste their output into a ticket, an email or a chat.

    Tunnel paths →

  • ●CharlotteHA partner missing
    • HA: only 1 of 2 FortiGates in the cluster – no failover (WARNING: FGDEMO00060002 is lost @ 2026/11/10 22:14:05)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system ha status
    HA Health Status: WARNING: FGDEMO00060002 is lost @ 2026/11/10 22:14:05
    How to fix

    Only one FortiGate is in the cluster, so there is no failover if it dies.

    1. Check the second unit is powered on.
    2. Check the HA heartbeat cable/port on both units.
    3. Confirm both units have the same HA group name, group ID and password.
    get system ha status
    diagnose sys ha history read
  • ●Chicagono tunnel to Boston, Philadelphia
    • No working IPsec tunnel to Boston, Philadelphia – every tunnel to that site is down
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn ipsec tunnel summary
    'to-boston' 203.0.113.25:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    'to-boston-b' 198.51.100.25:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    'to-philadelphia' 203.0.113.28:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    'to-philadelphia-b' 198.51.100.28:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    How to fix

    Some IPsec tunnels to other sites are down. If the aggregate's other member is up, traffic still flows – this is lost redundancy, not an outage. Backup tunnels whose phase 2 isn't set to auto-negotiate stay down until traffic needs them, so they look down here.

    1. Check the tunnel from this end (and the other site's end) – the WAN page's Tunnel paths tab shows both.
    2. For backup tunnels, enable auto-negotiate on their phase 2 (below) so they stay up.
    3. Tunnels that point at the wrong IP or a site that no longer exists: fix the remote gateway, or delete them.

    to-boston → 203.0.113.25
    to-boston-b → 198.51.100.25
    to-philadelphia → 203.0.113.28
    to-philadelphia-b → 198.51.100.28

    get vpn ipsec tunnel summary
    diagnose vpn tunnel list name to-boston
    diagnose vpn tunnel list name to-boston-b
    diagnose vpn tunnel list name to-philadelphia

    Tunnel paths →

  • ●Chicago4 tunnels down
    • 4 tunnels down: to-boston, to-boston-b, to-philadelphia, to-philadelphia-b
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn ipsec tunnel summary
    'to-boston' 203.0.113.25:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    'to-boston-b' 198.51.100.25:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    'to-philadelphia' 203.0.113.28:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    'to-philadelphia-b' 198.51.100.28:0  selectors(total,up): 1/0  rx(pkt,err): 0/0  tx(pkt,err): 0/0
    How to fix

    Some IPsec tunnels to other sites are down. If the aggregate's other member is up, traffic still flows – this is lost redundancy, not an outage. Backup tunnels whose phase 2 isn't set to auto-negotiate stay down until traffic needs them, so they look down here.

    1. Check the tunnel from this end (and the other site's end) – the WAN page's Tunnel paths tab shows both.
    2. For backup tunnels, enable auto-negotiate on their phase 2 (below) so they stay up.
    3. Tunnels that point at the wrong IP or a site that no longer exists: fix the remote gateway, or delete them.

    to-boston → 203.0.113.25
    to-boston-b → 198.51.100.25
    to-philadelphia → 203.0.113.28
    to-philadelphia-b → 198.51.100.28

    get vpn ipsec tunnel summary
    diagnose vpn tunnel list name to-boston
    diagnose vpn tunnel list name to-boston-b
    diagnose vpn tunnel list name to-philadelphia

    Tunnel paths →

  • ●ChicagoFortiLink split-interface with MCLAG
    • FortiLink aggregate fortilink still has fortilink-split-interface enabled while an MCLAG ICL is present. Split-interface keeps only one FortiLink path active – disable it for FortiLink MCLAG.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system interface
    edit "fortilink" … fortilink-split-interface (not disable)
    diagnose switch-controller switch-info mclag icl
    2 ICL trunk(s)
    How to fix

    FortiLink split-interface keeps only one uplink active toward the FortiGate. With an MCLAG ICL already formed, leave split-interface disabled so both switches forward on FortiLink.

    1. On the FortiLink aggregate (fortilink), set fortilink-split-interface disable.
    2. Confirm LACP active on the FortiLink aggregate.
    3. Re-check mclag icl and peer-consistency-check.
    show system interface
    
    config system interface
        edit "fortilink"
            set fortilink-split-interface disable
            set lacp-mode active
        next
    end
    
    diagnose switch-controller switch-info mclag icl
  • ●DenverIPsec errors climbing
    • 1 tunnel(s) with IPsec errors climbing since the last comparison: to-chicago (rx 41 / tx 0)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn ipsec tunnel summary
    'to-chicago' 203.0.113.11:0  selectors(total,up): 1/1  rx(pkt,err): 61080/41  tx(pkt,err): 63800/0
    How to fix

    IPsec error counters on these tunnels have climbed since the last comparison. These are the tunnel's own drop counts, not WAN health-check loss.

    1. Check the tunnel from both ends (below).
    2. A selector 2/1 often means one phase 2 is down – enable auto-negotiate, or the two ends' selectors disagree.
    3. Errors climbing: look at the phase-2 proposal and whether anti-replay is dropping because return traffic arrives on a different member.

    to-chicago: selectors 1/1, errors +rx 41 / +tx 0

    # safe to share (no keys):
    get vpn ipsec tunnel summary
    show vpn ipsec phase2-interface
    
    # The next commands print LIVE KEYS (tunnel encryption keys, pre-shared keys). Read them on the FortiGate –
    # never paste their output into a ticket, an email or a chat.
    diagnose vpn tunnel list name to-chicago

    Tunnel paths →

  • ●HoustonHA out of sync

    Mismatched checksum parts (all, root):

    • all
      • FGDEMO00020001=76 09 b4 06 79 89 51 ac ba b2 04 3f 3f 63 47 e5
      • FGDEMO00020002=20 bd d4 fc 9f e4 08 40 a0 a2 1e 64 d5 73 5b d1
    • root
      • FGDEMO00020001=9a 7d 73 f9 8c df 1c 77 da 5e af 60 27 f6 ad 9e
      • FGDEMO00020002=00 2b 13 8f 98 3b 24 7a 6b 91 35 05 a6 ff e2 2c
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system ha status
    Configuration Status:
        FGDEMO00020001(updated 2 seconds ago): in-sync
        FGDEMO00020002(updated 3 seconds ago): out-of-sync
    diagnose sys ha checksum cluster
    part        FGDEMO00020001    FGDEMO00020002    
    ------------------------------------------------
    all         76 09 b4 06 79 8  20 bd d4 fc 9f e  
    root        9a 7d 73 f9 8c d  00 2b 13 8f 98 3  
    How to fix

    The two FortiGates in the HA cluster have different configurations. If the primary fails, the secondary takes over with the wrong settings. A firmware mismatch between the units also causes this.

    1. Compare the checksums – the module that differs shows where.
    2. Make sure both units run the same firmware.
    3. Force a sync from the primary, wait a minute, and check again.

    Differs in: all, root – all = the overall checksum (follows the others); root = the root VDOM (firewall policies, addresses, routes, VPNs, SD-WAN). 'diagnose sys ha checksum show <part>' on both units lists a checksum per table – the table whose checksum differs is the object to compare.
    Mismatched parts only (matching parts omitted): part FGDEMO00020001 FGDEMO00020002 ------------------------------------------------all 76 09 b4 06 79 8 20 bd d4 fc 9f e root 9a 7d 73 f9 8c d 00 2b 13 8f 98 3

    diagnose sys ha checksum cluster
    diagnose sys ha checksum show root
    diagnose sys ha checksum recalculate
    get system ha status
    
    # still out of sync – force it from the secondary:
    execute ha manage <index> <admin>
    execute ha synchronize start
    
    # firmware on the secondary:
    get system status
  • ●HoustonFortiLink link down
    • FortiLink fortilink: 1 of 2 links down (port x2) – the switches are running on port x1 only
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system interface
    config system interface
        edit "fortilink"
            set fortilink enable
            set type aggregate
            set member "x1" "x2"
        next
    end
    get system interface physical
    ==[x1]
        status: up
        speed: 10000Mbps (Duplex: full)
    
    ==[x2]
        status: down
        speed: n/a
    How to fix

    The FortiGate reaches its switches over the fortilink aggregate. One member is down, so every switch, AP and wired device at the site now depends on a single cable/SFP – lose it and the whole LAN is cut off.

    1. x2 is disabled in the config – enable it (below) if it should be in use.
    2. Check the cable/fiber and the SFP on both ends (FortiGate port and the core switch's FortiLink port).
    3. Look at the transceiver readings – a dead or wrong-type SFP shows here.
    4. If the second link was never cabled, either cable it or remove it from the aggregate so it stops showing.
    get system interface physical x2
    get system interface transceiver
    diagnose netlink aggregate name fortilink
    
    # re-enable:
    config system interface
        edit "x2"
            set status up
        next
    end
  • ●HoustonHA single heartbeat
    • HA has 1 heartbeat device(s) (ha1) – a second hbdev keeps the cluster from splitting when one HA cable fails.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system ha
    config system ha
        set mode a-p
        set hbdev "ha1" 50
        set monitor "wan1" "wan2" "x1" "x2"
    end
    How to fix

    HA needs two heartbeat links so a single cable or port failure does not split the cluster (both units thinking they are primary).

    1. Add a second dedicated HA port (or a monitored VLAN) as hbdev.
    2. Use separate physical paths where you can.
    3. After the change, confirm both units list the same hbdev ports.
    show system ha
    
    config system ha
        set hbdev ha1 50
    end
  • ●Houstontrunk member not active

    Switch trunk member(s) not active – LACP redundancy is reduced.

    • FG-HOUB-SW1 trunk FGDEMO00020001: port28 not active (still up: port27)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    diagnose switch-controller switch-info trunk status
    FG-HOUB-SW1 trunk FGDEMO00020001: port28 not active (still up: port27)
    How to fix

    A FortiSwitch trunk (LACP) with a member not active has lost a redundant path. Spanning or failover still works on the remaining members, but one more failure cuts the link.

    1. Check the cable/SFP and the far-end port for each inactive member.
    2. Confirm LACP mode matches at both ends (active/passive).
    3. If the port was retired, remove it from the trunk so it stops showing.

    FG-HOUB-SW1 FGDEMO00020001: down port28

    diagnose switch-controller switch-info trunk status <switch>
  • ●New Yorkno OSPF neighbour
    • OSPF is set up on to-chicago but no neighbour is up there – routes to the far end come another way, or not at all
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get router info ospf neighbor
    OSPF process 0, VRF 0:
    Neighbor ID     Pri   State           Dead Time   Address         Interface
    How to fix

    OSPF is configured on this tunnel or aggregate and the tunnel is up, but no neighbour has formed. Usual causes: the aggregate has no running members (see that item), hellos leave on one member and return on another, or the two ends disagree on area, network type, timers or MTU.

    1. If this page also lists an aggregate that is not running, fix that first – OSPF cannot form over it.
    2. Compare the OSPF interface settings at both ends.
    3. Watch whether hellos arrive and leave (below).
    # on the PRIMARY unit (get system status | grep -i 'HA mode')
    # safe to share (no keys):
    get router info ospf neighbor
    get router info ospf interface
    diagnose sniffer packet any 'proto 89' 4 10
  • ●Seattleaggregate member not running
    • IPsec aggregate to-chicago: to-chicago is up but not in the running aggregate (1 of 2 members running) – including its first, preferred member
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system ipsec-aggregate
    config system ipsec-aggregate
        edit "to-chicago"
            set member "to-chicago" "to-chicago-b"
            set algorithm redundant
        next
    end
    diagnose sys ipsec-aggregate list
    bundle=to-chicago algo=redundant member=1
        tunnel=to-chicago-b weight=10
    get vpn ipsec tunnel summary
    'to-chicago' 203.0.113.11:0  selectors(total,up): 1/1  rx(pkt,err): 62128/0  tx(pkt,err): 64790/0
    'to-chicago-b' 198.51.100.11:0  selectors(total,up): 1/1  rx(pkt,err): 62278/0  tx(pkt,err): 64982/0
    How to fix

    These tunnels are up, but they are not in the running aggregate – so the FortiGate cannot send anything through them (it still receives, which is what one-way traffic looks like). The configuration is right; the running state is stale. It happens after a live aggregate or one of its tunnels is edited (members, weights, algorithm): a tunnel only joins again the next time it renegotiates.

    1. Renegotiate ONE of the missing tunnels (below). Only that tunnel drops, for a few seconds – start with one that carries nothing.
    2. Look at the running aggregate again: the tunnel should now be listed.
    3. Repeat for each tunnel still missing. If one does not come back, stop and open a Fortinet support case with the two outputs.
    4. After any later change to an aggregate or its tunnels: renegotiate and check again.

    Routes that point at this aggregate: 10.1.0.0/16, 10.2.0.0/16, 10.3.0.0/16, 10.4.0.0/16, 10.5.0.0/16, 10.6.0.0/16, 10.7.0.0/16, 10.8.0.0/16

    # on the PRIMARY unit (get system status | grep -i 'HA mode')
    diagnose vpn ike gateway flush name to-chicago
    
    # safe to share (no keys):
    diagnose sys ipsec-aggregate list
    get vpn ipsec tunnel summary

    Tunnel paths →

  • ▲BostonFG-BOSTON-SW1 slow ping
    • Collector ping to FG-BOSTON-SW1 (10.255.15.10): 263 ms, 33% loss
    Explain →
    How to fix

    The collector can ping FG-BOSTON-SW1, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲BostonBoston FortiGate slow ping
    • Collector ping to Boston FortiGate (10.250.0.15): 239 ms, 33% loss
    Explain →
    How to fix

    The collector can ping Boston FortiGate, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Bostonboston FortiGate (users) slow ping
    • Collector ping to boston FortiGate (users) (10.15.10.1): 292 ms, 33% loss
    Explain →
    How to fix

    The collector can ping boston FortiGate (users), but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Bostonboston servers gateway slow ping
    • Collector ping to boston servers gateway (10.15.30.1): 211 ms, 33% loss
    Explain →
    How to fix

    The collector can ping boston servers gateway, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Bostonboston printers gateway slow ping
    • Collector ping to boston printers gateway (10.15.40.1): 260 ms, 0% loss
    Explain →
    How to fix

    The collector can ping boston printers gateway, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Bostonboston cameras gateway slow ping
    • Collector ping to boston cameras gateway (10.15.50.1): 270 ms, 33% loss
    Explain →
    How to fix

    The collector can ping boston cameras gateway, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Bostonboston iot gateway slow ping
    • Collector ping to boston iot gateway (10.15.90.1): 263 ms, 33% loss
    Explain →
    How to fix

    The collector can ping boston iot gateway, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Bostonboston FortiLink slow ping
    • Collector ping to boston FortiLink (10.255.15.1): 284 ms, 33% loss
    Explain →
    How to fix

    The collector can ping boston FortiLink, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲BostonFG-BOSTON-AP1 slow ping
    • Collector ping to FG-BOSTON-AP1 (10.15.10.20): 229 ms, 33% loss
    Explain →
    How to fix

    The collector can ping FG-BOSTON-AP1, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲BostonFG-BOSTON-AP2 slow ping
    • Collector ping to FG-BOSTON-AP2 (10.15.10.21): 275 ms, 33% loss
    Explain →
    How to fix

    The collector can ping FG-BOSTON-AP2, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Bostonboston DNS 10.1.30.2 slow ping
    • Collector ping to boston DNS 10.1.30.2 (10.1.30.2): 272 ms, 0% loss
    Explain →
    How to fix

    The collector can ping boston DNS 10.1.30.2, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲Bostonboston DNS 10.1.30.3 slow ping
    • Collector ping to boston DNS 10.1.30.3 (10.1.30.3): 263 ms, 33% loss
    Explain →
    How to fix

    The collector can ping boston DNS 10.1.30.3, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ▲DenverExample Fiber gateway slow ping
    • Collector ping to Example Fiber gateway (203.0.113.17): 289 ms, 0% loss
    Explain →
    How to fix

    The collector can ping Example Fiber gateway, but latency or loss is high.

    1. Confirm the address still answers from the collector (ping).
    2. Check routing and firewall policy between the collector and the target.
    3. If this is an ISP gateway, check the circuit on Circuits & contacts.

    Ping →

  • ●Atlanta1 AP at 100 Mb
    • Access point(s) linked below 1 Gb – usually a bad cable or a 2-pair run: FG-ATLANTA-AP1 (100Base-TX on FG-ATLANTA-SW1 port3)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get wireless-controller wtp-status
    WTP: FP231F09010001
        name             : FG-ATLANTA-AP1
        sys name         : FG-ATLANTA-SW1
        port id          : port3
        MAU oper type    : 100Base-TX
    How to fix

    These APs negotiated 100 Mb instead of 1 Gb, which caps every Wi-Fi client behind them. Usually a damaged cable, a bad keystone/patch, or a run with only two pairs terminated.

    1. Swap the patch cables at both ends.
    2. If it stays at 100 Mb, test/re-terminate the cable run.
    3. Bounce the switch port to renegotiate after a fix.
    # The switch and port below are what the access point ITSELF announced (LLDP) - they were not read from the switch.
    # Confirm them before resetting power: the Find page, or the first command.
    diagnose switch-controller switch-info lldp neighbors-summary
    diagnose switch-controller switch-info port-stats FG-ATLANTA-SW1 port3
    execute switch-controller poe-reset FG-ATLANTA-SW1 port3
  • ●Chicago1 optic to check
    • FortiGate x2: Rx power warning (-9.62)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system interface transceiver / modules summary
    Interface x1 - SFP/SFP+, 10GBASE-SR
      Vendor Name  :           FORTINET
      Part No.     :           FN-TRAN-SFP+SR
      Serial No.   :           NLFG35644
    Interface x2 - SFP/SFP+, 10GBASE-SR
      Vendor Name  :           FORTINET
      Part No.     :           FN-TRAN-SFP+SR
      Serial No.   :           NLFG02130
    Optical       Temperature        Voltage         Tx Bias        Tx Power        Rx Power
    Interface     (Celsius)          (Volts)         (mA)           (dBm)           (dBm)
    ------------  ---------------    -------------   -------------  -------------   -------------
    x1            36.8               3.30            6.42           -2.31           -3.05
    x2            38.1               3.29            6.57           -2.44           -9.62 -
      ++ : high alarm, + : high warning, - : low warning, -- : low alarm, ? : suspect.
    How to fix

    Optics reporting alarms or warnings, no light on an inserted module, or low receive power. Low light usually means a dirty connector or a damaged/long fiber – it gets worse until the link drops. An inserted optic with no light on an unused port is just a spare – remove it to clear the alarm.

    1. Clean both ends of the fiber (and the optic) with a fiber cleaner; re-seat the optic.
    2. Still low: swap the patch cable, then the optic; check the fiber type matches the optic (SR = multimode, LR = single-mode).
    3. Unused port with an optic: remove it.

    FortiGate x2

    get system interface transceiver
    diagnose switch-controller switch-info modules summary <switch>
    diagnose switch-controller switch-info modules detail <switch> <port>
  • ●DenverExample Fiber slow
    • WAN Example Fiber (wan1) is slow on 1 of 1 health checks – worst: 280 ms, 1 ms jitter, 0% loss (Internet)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    diagnose sys sdwan health-check
    Health Check(Internet):
      Seq(1 wan1): state(alive), packet-loss(0.000%) latency(280.000), jitter(1.000), mos(4.300), bandwidth-up(400000), bandwidth-dw(400000), bandwidth-bi(800000) sla_map=0x0
    How to fix

    The Example Fiber connection is up but over its limits (latency 150 ms, jitter 30 ms, loss 2%). Calls and remote desktops suffer first.

    1. Check whether the link is simply full (bandwidth use on the port).
    2. If it persists and the link isn't full, open a ticket with the ISP.
    3. Cellular / satellite links have looser limits; to tune them, set "sla" on that ISP in the site file.
    diagnose sys sdwan health-check
    diagnose netlink interface list wan1

    Circuits & contacts →

  • ●Denver1 optic to check
    • FG-DENVER-SW1 port28: 10G-Base-SR INSERT, receive LOS – remove the optic if the port is unused
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system interface transceiver / modules summary
    FG-DENVER-SW1 port28: 10G-Base-SR INSERT, receive LOS – remove the optic if the port is unused
    How to fix

    Optics reporting alarms or warnings, no light on an inserted module, or low receive power. Low light usually means a dirty connector or a damaged/long fiber – it gets worse until the link drops. An inserted optic with no light on an unused port is just a spare – remove it to clear the alarm.

    1. Clean both ends of the fiber (and the optic) with a fiber cleaner; re-seat the optic.
    2. Still low: swap the patch cable, then the optic; check the fiber type matches the optic (SR = multimode, LR = single-mode).
    3. Unused port with an optic: remove it.

    FG-DENVER-SW1 port28

    get system interface transceiver
    diagnose switch-controller switch-info modules summary <switch>
    diagnose switch-controller switch-info modules detail <switch> <port>
  • ●Houston2 optics to check
    • FortiGate x2: Rx power ALARM (-40.0), Tx power ALARM (-40.0), temperature ALARM (73.5), Tx bias ALARM (0.0); FG-HOUB-SW1 port28: 10G-Base-SR INSERT, receive LOS – remove the optic if the port is unused
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system interface transceiver / modules summary
    Interface x1 - SFP/SFP+, 10GBASE-SR
      Vendor Name  :           FORTINET
      Part No.     :           FN-TRAN-SFP+SR
      Serial No.   :           NLFG97537
    Interface x2 - SFP/SFP+, 10GBASE-SR
      Vendor Name  :           FORTINET
      Part No.     :           FN-TRAN-SFP+SR
      Serial No.   :           NLFG33433
    Optical       Temperature        Voltage         Tx Bias        Tx Power        Rx Power
    Interface     (Celsius)          (Volts)         (mA)           (dBm)           (dBm)
    ------------  ---------------    -------------   -------------  -------------   -------------
    x1            41.3               3.31            6.10           -2.62           -3.48
    x2            73.5 ++            3.27            0.00 --        -40.00 --       -40.00 --
      ++ : high alarm, + : high warning, - : low warning, -- : low alarm, ? : suspect.
    
    FG-HOUB-SW1 port28: 10G-Base-SR INSERT, receive LOS – remove the optic if the port is unused
    How to fix

    Optics reporting alarms or warnings, no light on an inserted module, or low receive power. Low light usually means a dirty connector or a damaged/long fiber – it gets worse until the link drops. An inserted optic with no light on an unused port is just a spare – remove it to clear the alarm.

    1. Clean both ends of the fiber (and the optic) with a fiber cleaner; re-seat the optic.
    2. Still low: swap the patch cable, then the optic; check the fiber type matches the optic (SR = multimode, LR = single-mode).
    3. Unused port with an optic: remove it.

    FortiGate x2
    FG-HOUB-SW1 port28

    get system interface transceiver
    diagnose switch-controller switch-info modules summary <switch>
    diagnose switch-controller switch-info modules detail <switch> <port>
  • ●AtlantaVPN allow one way only
    • 1 IPsec tunnel interface(s) have an accept policy in one direction only: to-chicago (outbound only (LAN → tunnel); policy 9, 11, 12). Without a return allow, replies die on this FortiGate and counters look one-way.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show firewall policy
    edit "9"
    edit "11"
    diagnose vpn tunnel list
    name=to-chicago ver=2 serial=1 203.0.113.19:0->203.0.113.11:0 nexthop=203.0.113.17 tun_id=203.0.113.11 tun_id6=::203.0.113.11 dst_mtu=1500 dpd-link=on weight=1
    bound_if=5 lgwy=static/1 tun=intf mode=auto/1 encap=none/552 options[0228]=npu frag-rfc  run_state=0 role=primary accept_traffic=1 overlay_id=0
    
    proxyid_num=1 child_num=0 refcnt=5 ilast=0 olast=1 ad=/0
    stat: rxp=90000 txp=0 rxb=12780000 txb=0
    dpd: mode=on-demand on=1 status=ok idle=20000ms retry=3 count=0 seqno=223
    natt: mode=none draft=0 interval=0 remote_port=0
    fec: egress=0 ingress=0
    proxyid=to-chicago proto=0 sa=1 ref=2 serial=1 auto-negotiate
      src: 0:0.0.0.0-255.255.255.255:0
      dst: 0:0.0.0.0-255.255.255.255:0
      SA:  ref=3 options=18227 type=00 soft=0 mtu=1438 expire=9387/0B replaywin=2048
           seqno=1 esn=0 replaywin_lastseq=00015f90 qat=0 rekey=0 hash_search_len=1
      life: type=01 bytes=0/0 timeout=9387/43200
      dec: spi=24ac0205 esp=aes key=32 <removed>
           ah=sha256 key=32 <removed>
      enc: spi=c242fdcb esp=aes key=32 <removed>
           ah=sha256 key=32 <removed>
      dec:pkts/bytes=90000/12780000, enc:pkts/bytes=0/0
      npu_flag=03 npu_rgwy=203.0.113.11 npu_lgwy=203.0.113.19 npu_selid=0 dec_npuid=1 enc_npuid=1
    ------------------------------------------------------
    How to fix

    An accept firewall policy uses this IPsec tunnel interface in only one direction. Replies that need the opposite allow are dropped here, which shows up as one-way tunnel counters. Dial-up and DECOM tunnels are not listed.

    1. Open the listed policy and note its source, destination and services.
    2. Add a matching accept in the reverse direction (or a zone pair that covers both).
    3. Do not use `any` for WAN-facing source; keep the path as narrow as the outbound allow.
    4. Suggested CLI below is a sketch – replace the placeholders before pasting.
    # to-chicago: add an accept with srcintf "to-chicago" toward the LAN (mirror the outbound allow)
    config firewall policy
        edit 0
            set name "return-to-chicago"
            set srcintf "to-chicago"
            set dstintf "<lan-or-zone>"
            set srcaddr "<remote-nets>"
            set dstaddr "<local-nets>"
            set action accept
            set schedule always
            set service ALL
            set logtraffic all
        next
    end
  • ●Atlanta1 stale AP entry
    • 1 AP(s) configured on the FortiGate that have never connected: FG-ATLANTA-APNEW
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get wireless-controller wtp-status
    WTP: FP231F09900001
        wtp-id           : FP231F09900001
        name             : FG-ATLANTA-APNEW
        connection-state : Disconnected
        join-time        : N/A
    name: to-chicago
    name: to-chicago-b
    How to fix

    These APs are configured on the FortiGate but have never connected. They clutter the wireless controller and hide real problems. Delete the ones that no longer exist (or acknowledge them if they're still to be installed).

    1. Confirm they aren't APs waiting to be installed.
    2. Delete them on the FortiGate (below).
    config wireless-controller wtp
        delete "FG-ATLANTA-APNEW"
    end
  • ●Atlantafirmware below baseline
    • 1 device(s) below the firmware baseline: FG-ATLANTA-AP1 7.2.2 (baseline 7.4.3)
    Explain →
    How to fix

    These devices run firmware older than the baseline your security team set (config.json → security → firmware_baseline). Older firmware usually means known, fixed vulnerabilities are still present.

    1. Check the upgrade path for each model (Fortinet upgrade path tool) and the release notes.
    2. Plan a maintenance window (maintenance.py add …) and upgrade – FortiGate first, then switches and APs from the FortiGate.
    3. If a device can't be upgraded (end of support), replace it or record the exception with the security team.

    FortiAP FG-ATLANTA-AP1 (FortiAP 231F): 7.2.2 → at least 7.4.3

    get system status
    diagnose switch-controller switch-info status
    get wireless-controller wtp-status
  • ●Bostonaggregate-member flag
    • in an aggregate with aggregate-member disabled: to-chicago (chicago); to-dallas (chicago)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show vpn ipsec phase1-interface
    config vpn ipsec phase1-interface
        edit "to-chicago"
            set interface "wan1"
            set remote-gw 203.0.113.11
        next
    end
    
    config vpn ipsec phase1-interface
        edit "to-dallas"
            set interface "wan1"
            set remote-gw 203.0.113.18
        next
    end
    show system ipsec-aggregate
    config system ipsec-aggregate
        edit "chicago"
            set member "to-chicago" "to-dallas"
            set algorithm redundant
        next
    end
    How to fix

    A tunnel listed in an IPsec aggregate should have aggregate-member enable; a tunnel with that flag and no aggregate is left out of the bundle. Either way the FortiGate may send on a member the other end does not expect.

    1. Enable aggregate-member on every tunnel that is in an aggregate.
    2. Clear the flag on a tunnel that is not in any aggregate, or add it to the aggregate at both ends.
    config vpn ipsec phase1-interface
        edit "to-chicago"
            set aggregate-member enable
        next
        edit "to-dallas"
            set aggregate-member enable
        next
    end
    show system ipsec-aggregate
  • ●Bostonaggregate not paired
    • IPsec aggregate chicago: members do not all end at one known site, so the order, mode and missing-member checks did not run (peers: Chicago, Dallas)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system ipsec-aggregate
    config system ipsec-aggregate
        edit "chicago"
            set member "to-chicago" "to-dallas"
            set algorithm redundant
        next
    end
    How to fix

    The both-ends checks (order, mode, missing member) only run when every member of an aggregate ends at one known site. When they do not, or when both members leave on the same local WAN, those checks are skipped or treat the members as two paths by the far circuit – this note says so.

    1. If the members should all go to one site, fix the remote gateways so they map.
    2. If two members share a WAN on purpose, the path order is still by the far end's circuit.
    # safe to share (no keys):
    show system ipsec-aggregate
    show vpn ipsec phase1-interface
    get vpn ipsec tunnel summary
  • ●Chicagoclear-text admin
    • Unencrypted admin access (http/telnet) enabled: users (http)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system interface
    config system interface
        edit "users"
            set ip 10.1.10.1 255.255.255.0
            set allowaccess ping http
        next
    end
    How to fix

    http/telnet send admin passwords unencrypted. Use https (and ssh) only for the admin GUI.

    1. Confirm nobody manages this FortiGate over http or telnet on these interfaces.
    2. Remove http and telnet from allowaccess; keep https (and ssh if you use it) – below keeps everything else.
    3. Also check Best Practices for strong-crypto, admin-https-ssl-versions, and admin-server-cert.
    config system interface
        edit "users"
            set allowaccess ping
        next
    end
  • ●ChicagoIPsec settings differ at the two ends
    • Chicago's to-seattle-b ↔ Seattle's to-chicago-b: phase 2 auto-negotiate: enable vs (default)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show vpn ipsec phase1-interface
    config vpn ipsec phase1-interface
        edit "to-seattle-b"
            set remote-gw 198.51.100.21
            set ike-version 2
            set proposal aes256-sha256
            set dhgrp 14
        next
    end
    
    config vpn ipsec phase1-interface
        edit "to-chicago-b"
            set remote-gw 198.51.100.11
            set ike-version 2
            set proposal aes256-sha256
            set dhgrp 14
        next
    end
    How to fix

    The two ends of this path disagree on phase 1 or phase 2 (IKE version, proposal, DH group, dead-peer detection, key life, auto-negotiate, or the phase-2 selectors). `show` omits a setting left at the FortiOS default; a named selector cannot be expanded to its subnets here. A value written on one end and absent on the other is still a mismatch. When the tunnel is one-way and the aggregates already agree, this is the next place to look.

    1. Compare the settings below at both ends.
    2. Change both ends together in a maintenance window – a mismatch takes the tunnel down.
    3. Named address objects are not collected; check those by hand if a selector is a name on one end.
    # this end:
    show vpn ipsec phase1-interface | grep -f to-seattle-b
    show vpn ipsec phase2-interface | grep -f to-seattle-b
    
    # on Seattle:
    show vpn ipsec phase1-interface | grep -f to-chicago-b
    show vpn ipsec phase2-interface | grep -f to-chicago-b
  • ●ChicagoIPsec settings differ at the two ends
    • Chicago's to-losangeles ↔ Los Angeles's to-chicago: phase 1 ike-version: 2 vs (default); phase 1 proposal: aes256-sha256 vs aes128-sha1; phase 1 dhgrp: 14 vs 5
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show vpn ipsec phase1-interface
    config vpn ipsec phase1-interface
        edit "to-losangeles"
            set remote-gw 203.0.113.23
            set ike-version 2
            set proposal aes256-sha256
            set dhgrp 14
        next
    end
    
    config vpn ipsec phase1-interface
        edit "to-chicago"
            set remote-gw 203.0.113.11
            set proposal aes128-sha1
            set dhgrp 5
        next
    end
    How to fix

    The two ends of this path disagree on phase 1 or phase 2 (IKE version, proposal, DH group, dead-peer detection, key life, auto-negotiate, or the phase-2 selectors). `show` omits a setting left at the FortiOS default; a named selector cannot be expanded to its subnets here. A value written on one end and absent on the other is still a mismatch. When the tunnel is one-way and the aggregates already agree, this is the next place to look.

    1. Compare the settings below at both ends.
    2. Change both ends together in a maintenance window – a mismatch takes the tunnel down.
    3. Named address objects are not collected; check those by hand if a selector is a name on one end.
    # this end:
    show vpn ipsec phase1-interface | grep -f to-losangeles
    show vpn ipsec phase2-interface | grep -f to-losangeles
    
    # on Los Angeles:
    show vpn ipsec phase1-interface | grep -f to-chicago
    show vpn ipsec phase2-interface | grep -f to-chicago
  • ●Dallasold certificate
    • Expired certificate still installed: retired-vpn (expired Sep 04 2026)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn certificate local details
    retired-vpn: valid to 2026-09-04
    How to fix

    The certificate 'retired-vpn' expired on 2026-09-04 and is still installed. Nothing can have been working with it since then, so it's a leftover – but it clutters the certificate list and could be picked by mistake.

    1. Check nothing references it (below) – SSL VPN, admin GUI, IPsec, SSL inspection.
    2. Delete it. FortiOS refuses to delete a certificate that's still in use, so this is safe to try.
    3. Or, to keep it, acknowledge it in the site file: "acknowledged": {"old_cert": {"<name>": "why"}}.
    show vpn ssl settings | grep -f cert
    show system global | grep -f cert
    config vpn certificate local
        delete "retired-vpn"
    end
  • ●Dallasfirmware below baseline
    • 2 device(s) below the firmware baseline: FG-DALLAS 7.2.4 (baseline 7.4.8), FG-DALLAS-SW1 7.0.5 (baseline 7.4.5)
    Explain →
    How to fix

    These devices run firmware older than the baseline your security team set (config.json → security → firmware_baseline). Older firmware usually means known, fixed vulnerabilities are still present.

    1. Check the upgrade path for each model (Fortinet upgrade path tool) and the release notes.
    2. Plan a maintenance window (maintenance.py add …) and upgrade – FortiGate first, then switches and APs from the FortiGate.
    3. If a device can't be upgraded (end of support), replace it or record the exception with the security team.

    FortiGate FG-DALLAS (FortiGate 90G): 7.2.4 → at least 7.4.8
    FortiSwitch FG-DALLAS-SW1 (FortiSwitch 124F-FPOE): 7.0.5 → at least 7.4.5

    get system status
    diagnose switch-controller switch-info status
    get wireless-controller wtp-status
  • ●Los Angelesweak IPsec settings
    • 1 tunnel(s) with legacy crypto: to-chicago (IKEv1, SHA1/MD5/3DES offered, DH 5)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show vpn ipsec phase1-interface (no 'set ike-version 2' line = IKEv1)
    config vpn ipsec phase1-interface
        edit "to-chicago"
            set remote-gw 203.0.113.11
            set proposal aes128-sha1
            set dhgrp 5
        next
    end
    How to fix

    These tunnels use legacy crypto: IKEv1, SHA1/MD5/3DES proposals, or Diffie-Hellman groups below 14. They work, but they're weaker and IKEv1 is being phased out. Change both ends together in a maintenance window – a mismatch takes the tunnel down.

    1. Pick one tunnel pair and change both ends at the same time.
    2. Use IKEv2, aes256-sha256 / aes256gcm, DH 19/20/21 (keep 14 only if a peer needs it).
    3. Check it re-establishes, then move on to the next pair.

    to-chicago: IKEv1, SHA1/MD5/3DES offered, DH 5

    config vpn ipsec phase1-interface
        edit "<tunnel>"
            set ike-version 2
            set proposal aes256-sha256 aes256gcm-prfsha384
            set dhgrp 21 20 19
        next
    end
  • ●Los AngelesIPsec settings differ at the two ends
    • Chicago's to-losangeles ↔ Los Angeles's to-chicago: phase 1 ike-version: 2 vs (default); phase 1 proposal: aes256-sha256 vs aes128-sha1; phase 1 dhgrp: 14 vs 5
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show vpn ipsec phase1-interface
    config vpn ipsec phase1-interface
        edit "to-losangeles"
            set remote-gw 203.0.113.23
            set ike-version 2
            set proposal aes256-sha256
            set dhgrp 14
        next
    end
    
    config vpn ipsec phase1-interface
        edit "to-chicago"
            set remote-gw 203.0.113.11
            set proposal aes128-sha1
            set dhgrp 5
        next
    end
    How to fix

    The two ends of this path disagree on phase 1 or phase 2 (IKE version, proposal, DH group, dead-peer detection, key life, auto-negotiate, or the phase-2 selectors). `show` omits a setting left at the FortiOS default; a named selector cannot be expanded to its subnets here. A value written on one end and absent on the other is still a mismatch. When the tunnel is one-way and the aggregates already agree, this is the next place to look.

    1. Compare the settings below at both ends.
    2. Change both ends together in a maintenance window – a mismatch takes the tunnel down.
    3. Named address objects are not collected; check those by hand if a selector is a name on one end.
    # this end:
    show vpn ipsec phase1-interface | grep -f to-chicago
    show vpn ipsec phase2-interface | grep -f to-chicago
    
    # on Chicago:
    show vpn ipsec phase1-interface | grep -f to-losangeles
    show vpn ipsec phase2-interface | grep -f to-losangeles
  • ●New Yorkno blackhole routes
    • No blackhole route for 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16. When a VPN route disappears (tunnel or OSPF down), traffic for that remote network follows the default route to the internet instead of being dropped – and sessions can stick to the wrong path after the tunnel recovers.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get router info routing-table all
    S*      0.0.0.0/0 [1/0] via 203.0.113.17, wan1, [1/0]
    How to fix

    Fortinet's VPN best practice: a blackhole route for each private range at a high distance (254). If a tunnel or OSPF drops, the more specific VPN route disappears and the blackhole catches the traffic – instead of the default route sending internal traffic to the internet (and sessions sticking to that path after recovery).

    1. Add the routes below (distance 254, so any real route wins).
    2. Make sure no local network is only reachable through the default route.
    3. Check: get router info routing-table all shows the blackholes with [254/0].
    config router static
        edit 0
            set dst 10.0.0.0/8
            set blackhole enable
            set distance 254
            set comment "RFC1918 blackhole"
        next
        edit 0
            set dst 172.16.0.0/12
            set blackhole enable
            set distance 254
            set comment "RFC1918 blackhole"
        next
        edit 0
            set dst 192.168.0.0/16
            set blackhole enable
            set distance 254
            set comment "RFC1918 blackhole"
        next
    end
  • ●Seattleauto-negotiate off
    • 1 aggregate member(s) with phase 2 auto-negotiate off – they stay down until traffic needs them, so a failover waits for the tunnel to negotiate: to-chicago-b
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show vpn ipsec phase2-interface (no 'set auto-negotiate enable' line = off)
    config vpn ipsec phase2-interface
        edit "to-chicago-b"
            set phase1name "to-chicago-b"
        next
    end
    How to fix

    These backup tunnels only negotiate phase 2 when traffic needs them. When the primary fails, traffic waits for the backup to come up (and a NATed end like Azure may never start it), and this page can't tell an idle backup from a broken one. Enable auto-negotiate at both ends.

    1. Paste the block below on this FortiGate (it only changes the listed phase 2s).
    2. Do the same at the other end of each tunnel.
    3. Check: get vpn ipsec tunnel summary – each should show selectors 1/1.
    config vpn ipsec phase2-interface
        edit "to-chicago-b"
            set auto-negotiate enable
        next
    end
    get vpn ipsec tunnel summary
  • ●Seattleaggregate-member flag
    • in an aggregate with aggregate-member disabled: to-chicago (to-chicago); to-chicago-b (to-chicago)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show vpn ipsec phase1-interface
    config vpn ipsec phase1-interface
        edit "to-chicago"
            set interface "wan1"
            set remote-gw 203.0.113.11
        next
    end
    
    config vpn ipsec phase1-interface
        edit "to-chicago-b"
            set interface "wan2"
            set remote-gw 198.51.100.11
        next
    end
    show system ipsec-aggregate
    config system ipsec-aggregate
        edit "to-chicago"
            set member "to-chicago" "to-chicago-b"
            set algorithm redundant
        next
    end
    How to fix

    A tunnel listed in an IPsec aggregate should have aggregate-member enable; a tunnel with that flag and no aggregate is left out of the bundle. Either way the FortiGate may send on a member the other end does not expect.

    1. Enable aggregate-member on every tunnel that is in an aggregate.
    2. Clear the flag on a tunnel that is not in any aggregate, or add it to the aggregate at both ends.
    config vpn ipsec phase1-interface
        edit "to-chicago"
            set aggregate-member enable
        next
        edit "to-chicago-b"
            set aggregate-member enable
        next
    end
    show system ipsec-aggregate
  • ●Seattleweights on a redundant aggregate
    • IPsec aggregate to-chicago is 'redundant' and has weights set (to-chicago-b 10). In redundant mode the first member that is up carries the traffic – the order of 'set member' decides, not the weight. First member: to-chicago
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system ipsec-aggregate
    config system ipsec-aggregate
        edit "to-chicago"
            set member "to-chicago" "to-chicago-b"
            set algorithm redundant
        next
    end
    show vpn ipsec phase1-interface
    config vpn ipsec phase1-interface
        edit "to-chicago-b"
            set aggregate-weight 10
        next
    end
    How to fix

    Fortinet documents 'redundant' as 'use the first tunnel that is up'; aggregate-weight belongs to weighted-round-robin. Weights here suggest the heaviest tunnel was expected to be preferred – what decides is the order of the members, and it has to be the same path order at both ends.

    1. Put the tunnel on the main circuit first in 'set member' – at both ends, in the same window.
    2. Set the weights back to 1, so the configuration says what it does.
    3. Renegotiate the tunnels and check the running aggregate – editing a live aggregate can leave tunnels out of it.
    # on the PRIMARY unit (get system status | grep -i 'HA mode')
    config system ipsec-aggregate
        edit "to-chicago"
            set member "to-chicago" "to-chicago-b"     # current order – main circuit first
        next
    end
    config vpn ipsec phase1-interface
        edit "to-chicago-b"
            set aggregate-weight 1
        next
    end
    diagnose vpn ike gateway flush name to-chicago
    diagnose vpn ike gateway flush name to-chicago-b
    diagnose sys ipsec-aggregate list
  • ●Seattleaggregate not paired
    • IPsec aggregate to-chicago: Chicago has no IPsec aggregate whose members all end here, so the order, mode and missing-member checks did not run
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system ipsec-aggregate
    config system ipsec-aggregate
        edit "to-chicago"
            set member "to-chicago" "to-chicago-b"
            set algorithm redundant
        next
    end
    How to fix

    The both-ends checks (order, mode, missing member) only run when every member of an aggregate ends at one known site. When they do not, or when both members leave on the same local WAN, those checks are skipped or treat the members as two paths by the far circuit – this note says so.

    1. If the members should all go to one site, fix the remote gateways so they map.
    2. If two members share a WAN on purpose, the path order is still by the far end's circuit.
    # safe to share (no keys):
    show system ipsec-aggregate
    show vpn ipsec phase1-interface
    get vpn ipsec tunnel summary
  • ●SeattleIPsec settings differ at the two ends
    • Chicago's to-seattle-b ↔ Seattle's to-chicago-b: phase 2 auto-negotiate: enable vs (default)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show vpn ipsec phase1-interface
    config vpn ipsec phase1-interface
        edit "to-seattle-b"
            set remote-gw 198.51.100.21
            set ike-version 2
            set proposal aes256-sha256
            set dhgrp 14
        next
    end
    
    config vpn ipsec phase1-interface
        edit "to-chicago-b"
            set remote-gw 198.51.100.11
            set ike-version 2
            set proposal aes256-sha256
            set dhgrp 14
        next
    end
    How to fix

    The two ends of this path disagree on phase 1 or phase 2 (IKE version, proposal, DH group, dead-peer detection, key life, auto-negotiate, or the phase-2 selectors). `show` omits a setting left at the FortiOS default; a named selector cannot be expanded to its subnets here. A value written on one end and absent on the other is still a mismatch. When the tunnel is one-way and the aggregates already agree, this is the next place to look.

    1. Compare the settings below at both ends.
    2. Change both ends together in a maintenance window – a mismatch takes the tunnel down.
    3. Named address objects are not collected; check those by hand if a selector is a name on one end.
    # this end:
    show vpn ipsec phase1-interface | grep -f to-chicago-b
    show vpn ipsec phase2-interface | grep -f to-chicago-b
    
    # on Chicago:
    show vpn ipsec phase1-interface | grep -f to-seattle-b
    show vpn ipsec phase2-interface | grep -f to-seattle-b
  • ✕Philadelphiadata 50 h old
    • Data is 50 hours old – last successful collection Nov 20, 2:16 PM
    Explain →
    How to fix

    No successful collection for over a day – everything on this page may be out of date.

    1. Check the scheduled task ran (Task Scheduler → History of the hourly task).
    2. Read the latest log in C:\netmap\logs – a failed site's line ends with a diagnosis.
    3. Run the hourly task now and read the log again (manual runs go through the task – only its account can read the key).
    # on NET-DEMO01 (elevated PowerShell):
    Start-ScheduledTask "Vantage hourly"
    Get-Content (Get-ChildItem C:\netmap\logs\daily-*.log | Sort-Object LastWriteTime | Select-Object -Last 1) -Encoding UTF8 -Tail 30
  • ▲Miamidata 30 h old
    • Data is 30 hours old – last successful collection Nov 21, 10:16 AM
    Explain →
    How to fix

    No successful collection for over a day – everything on this page may be out of date.

    1. Check the scheduled task ran (Task Scheduler → History of the hourly task).
    2. Read the latest log in C:\netmap\logs – a failed site's line ends with a diagnosis.
    3. Run the hourly task now and read the log again (manual runs go through the task – only its account can read the key).
    # on NET-DEMO01 (elevated PowerShell):
    Start-ScheduledTask "Vantage hourly"
    Get-Content (Get-ChildItem C:\netmap\logs\daily-*.log | Sort-Object LastWriteTime | Select-Object -Last 1) -Encoding UTF8 -Tail 30
  • ●Azure EastLicence 3 days
    • Licence: FortiGate-VM licence expires in 3 days (Nov 25 2026)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system status
    FortiGate-VM licence: 2026-11-25
    How to fix

    The FortiGate-VM licence expires 2026-11-25 – after that the VM stops passing traffic.

    1. Renew through your Fortinet partner / FortiCare (support.fortinet.com → Assets) with the serial number.
    2. After the renewal is applied, the FortiGate picks it up on its next update check (or update it manually).
    3. If it's not going to be renewed, acknowledge it in the site file so it stops flagging.
    get system status
    diagnose autoupdate versions | grep -A2 -i expiry
    execute update-now   # admin account, after renewal
  • ●Charlottecircuit contract 74 days
    • Circuit contract: Metro Cable (ISP2) ends 2/4/2027 (74 days left)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system interface
    wan2: CT-ISP2-02042024-02042027
    How to fix

    The contract for Metro Cable (ISP2) ends on 2/4/2027. Renewing, renegotiating or moving to another provider all take time, and many contracts need written notice well before the end.

    1. Check the contract for the notice period and whether it renews by itself (the site's Library has the facts on file).
    2. Decide with the business: renew, renegotiate or replace the circuit.
    3. When the new term is signed, change the dates in the tag on the FortiGate interface comment.
    4. If nothing will be done, acknowledge it in the site file so it stops flagging.
    # the tag sits in the comment of the circuit's WAN interface (admin account):
    show system interface | grep -f CT-ISP
    
    config system interface
        edit "<wan interface>"
            set description "<what is there now> CT-ISP1-<MMDDYYYY>-<MMDDYYYY>"
        next
    end
  • ●Chicagohealth check misses a WAN
    • SD-WAN health check(s) do not cover wan4. Those links are not probed, so a silent ISP failure on them is not seen.
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system sdwan
            edit "virtual-wan-link"
            edit 1
                set interface "wan1"
    How to fix

    Each SD-WAN internet member should be in a health check so a quiet-but-dead ISP is detected. Link-state alone misses brownouts and upstream failures.

    1. Add a health check (or extend members=) so every WAN/SD-WAN member is probed.
    2. Use a public probe target reachable on those WANs.
    3. Confirm with diagnose sys sdwan health-check that each member shows alive or dead.

    Not covered: wan4

    show system sdwan
    diagnose sys sdwan health-check
    
    # example – probe every member:
    config system sdwan
        config health-check
            edit "Internet"
                set server "8.8.8.8"
                set members 1
            next
        end
    end
  • ●ChicagoCertificate 15 days
    • Certificate: Northline-Web expires in 15 days (Dec 07 2026)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn certificate local details
    Northline-Web: 2026-12-07
    How to fix

    The certificate 'Northline-Web' on this FortiGate expires 2026-12-07. Anything using it – SSL VPN, the admin GUI, an IPsec tunnel with certificate authentication – fails after that date.

    1. Find what uses it (below).
    2. Issue a new certificate from your CA, import it on the FortiGate.
    3. Point the services at the new certificate, then delete the old one.
    get vpn certificate local details Northline-Web
    show vpn ssl settings
    show system global | grep -f cert
    show vpn ipsec phase1-interface | grep -f certificate
  • ●ChicagoFortiGuard 33 days
    • FortiGuard: AV Engine, Virus Definitions, Extended set, Mobile Malware Definitions +2 more expires in 33 days (Dec 25 2026)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    diagnose autoupdate versions
    AV Engine, Virus Definitions, Extended set, Mobile Malware Definitions +2 more: 2026-12-25
    How to fix

    FortiGuard service(s) AV Engine, Virus Definitions, Extended set, Mobile Malware Definitions +2 more expire 2026-12-25. After that the FortiGate stops receiving updates for them (signatures, web filtering, etc.); traffic keeps flowing but protection goes stale.

    1. Renew through your Fortinet partner / FortiCare (support.fortinet.com → Assets) with the serial number.
    2. After the renewal is applied, the FortiGate picks it up on its next update check (or update it manually).
    3. If it's not going to be renewed, acknowledge it in the site file so it stops flagging.
    get system status
    diagnose autoupdate versions | grep -A2 -i expiry
    execute update-now   # admin account, after renewal
  • ●DallasCertificate 3 days
    • Certificate: northline-vpn expires in 3 days (Nov 25 2026)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn certificate local details
    northline-vpn: 2026-11-25
    How to fix

    The certificate 'northline-vpn' on this FortiGate expires 2026-11-25. Anything using it – SSL VPN, the admin GUI, an IPsec tunnel with certificate authentication – fails after that date.

    1. Find what uses it (below).
    2. Issue a new certificate from your CA, import it on the FortiGate.
    3. Point the services at the new certificate, then delete the old one.
    get vpn certificate local details northline-vpn
    show vpn ssl settings
    show system global | grep -f cert
    show vpn ipsec phase1-interface | grep -f certificate
  • ●Denvercollector
    • 1 collector command(s) failed: get router info ospf neighbor
    Explain →
    How to fix

    Some read-only commands failed on this FortiGate, so parts of the page may be missing. Usually a command that doesn't exist on this model/firmware, or isn't allowed for the netmap account.

    1. Look at the failed commands below; run one by hand as netmap to see the error.
    2. If it doesn't apply to this model, it can be skipped – send the error to whoever maintains Vantage.
    get router info ospf neighbor
  • ●Houstoncollected from the secondary
    • The collector is logged in to the secondary HA unit (FG-HOUB). Tunnel, routing and aggregate state on this page is that unit's – not what carries the traffic. Primary: FG-HOUSTON
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get system status
    Hostname: FG-HOUB
    Current HA mode: a-p, backup
    How to fix

    In a FortiGate HA pair one unit forwards the traffic (the primary). Tunnels, routes and aggregates are live there; the secondary's copies can look different, and troubleshooting from it gives wrong answers.

    1. Point the collector at the address that follows the primary, not at this unit's own management address (config.json → sites → host).
    2. When you work on the cluster by hand, check which unit you are on first (below).
    get system status | grep -i 'HA mode'
    get system ha status
  • ●HoustonCertificate 57 days
    • Certificate: Houston-MES expires in 57 days (Jan 18 2027)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn certificate local details
    Houston-MES: 2027-01-18
    How to fix

    The certificate 'Houston-MES' on this FortiGate expires 2027-01-18. Anything using it – SSL VPN, the admin GUI, an IPsec tunnel with certificate authentication – fails after that date.

    1. Find what uses it (below).
    2. Issue a new certificate from your CA, import it on the FortiGate.
    3. Point the services at the new certificate, then delete the old one.
    get vpn certificate local details Houston-MES
    show vpn ssl settings
    show system global | grep -f cert
    show vpn ipsec phase1-interface | grep -f certificate
  • ●Miamicircuit contract 38 days
    • Circuit contract: Example Fiber ends 12/30/2026 (38 days left)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    Library circuits.csv
    Example Fiber: contract_end 2026-12-30
    How to fix

    The contract for Example Fiber ends on 12/30/2026. Renewing, renegotiating or moving to another provider all take time, and many contracts need written notice well before the end.

    1. Check the contract for the notice period and whether it renews by itself (the site's Library has the facts on file).
    2. Decide with the business: renew, renegotiate or replace the circuit.
    3. When the new term is signed, change the dates in the tag on the FortiGate interface comment (or in the Library's circuits.csv).
    4. If nothing will be done, acknowledge it in the site file so it stops flagging.
    # the tag sits in the comment of the circuit's WAN interface (admin account):
    show system interface | grep -f CT-ISP
    
    config system interface
        edit "<wan interface>"
            set description "<what is there now> CT-ISP1-<MMDDYYYY>-<MMDDYYYY>"
        next
    end
  • ●Minneapoliscircuit contract ended
    • Circuit contract: Metro Cable (ISP2) ended 10/30/2026 (ended 23 days ago)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system interface
    wan2: CT-ISP2-10302021-10302026
    How to fix

    The contract for Metro Cable (ISP2) ended on 10/30/2026. The line may now be month to month, at a different price, or due to be switched off.

    1. Check the contract for the notice period and whether it renews by itself (the site's Library has the facts on file).
    2. Decide with the business: renew, renegotiate or replace the circuit.
    3. When the new term is signed, change the dates in the tag on the FortiGate interface comment.
    4. If nothing will be done, acknowledge it in the site file so it stops flagging.
    # the tag sits in the comment of the circuit's WAN interface (admin account):
    show system interface | grep -f CT-ISP
    
    config system interface
        edit "<wan interface>"
            set description "<what is there now> CT-ISP1-<MMDDYYYY>-<MMDDYYYY>"
        next
    end
  • ●Philadelphiahealth check
    • SD-WAN health check 'Internet' can't reach its target on any WAN – check its server setting
    Explain →
    How to fix

    The SD-WAN health check 'Internet' probes 198.51.100.53 203.0.113.53 over the internet links, and that address doesn't answer on any of them. A private (10.x) address can only be reached over a tunnel, so this check can never pass. Any SD-WAN rule using it is effectively switched off.

    1. If nothing depends on it: delete the SD-WAN rule that uses it, then the health check.
    2. If it's meant to watch a server across the VPN: make the IPsec tunnels its members instead of the WANs.
    show system sdwan
    
    config system sdwan
        config health-check
            delete "Internet"
        end
    end
  • ●Seattlehealth check
    • SD-WAN health check 'Monitor' can't reach its target on any WAN – check its server setting
    Explain →
    How to fix

    The SD-WAN health check 'Monitor' probes 10.1.30.10 over the internet links, and that address doesn't answer on any of them. A private (10.x) address can only be reached over a tunnel, so this check can never pass. Any SD-WAN rule using it is effectively switched off.

    1. If nothing depends on it: delete the SD-WAN rule that uses it, then the health check.
    2. If it's meant to watch a server across the VPN: make the IPsec tunnels its members instead of the WANs.
    show system sdwan
    
    config system sdwan
        config health-check
            delete "Monitor"
        end
    end
  • ●Seattlehealth check probes a private address
    • SD-WAN health check(s) probe a private address over WAN members – that target is not on the internet, so the check stays dead and any rule that uses it never fails over: Monitor → 10.1.30.10
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    show system sdwan
    Monitor: server 10.1.30.10
    How to fix

    An SD-WAN health check that probes a private address over WAN members can never succeed on the internet path. Rules that depend on it never see the link as up.

    1. Point the health check at a public address (or a well-known DNS/HTTP target).
    2. Or move the check onto IPsec tunnel members if it is meant to watch a remote private server.
    3. Remove the check if nothing uses it.
    show system sdwan
    diagnose sys sdwan health-check
  • ●SeattleCertificate 21 days
    • Certificate: Northline-Mgmt expires in 21 days (Dec 13 2026)
    Explain →
    Evidence

    Exact lines from the FortiGate output collected for this site – check them against the live device.

    get vpn certificate local details
    Northline-Mgmt: 2026-12-13
    How to fix

    The certificate 'Northline-Mgmt' on this FortiGate expires 2026-12-13. Anything using it – SSL VPN, the admin GUI, an IPsec tunnel with certificate authentication – fails after that date.

    1. Find what uses it (below).
    2. Issue a new certificate from your CA, import it on the FortiGate.
    3. Point the services at the new certificate, then delete the old one.
    get vpn certificate local details Northline-Mgmt
    show vpn ssl settings
    show system global | grep -f cert
    show vpn ipsec phase1-interface | grep -f certificate

Grouped by what is affected; within a group, what is down comes before what is degraded, and notes last. Groups start closed. Click a group heading to open it; Expand all / Collapse all do every group. Each item opens with the evidence from the FortiGate and the steps to fix it. Open is the default: muted, hidden, accepted and won't-fix findings stay in the list and do not change a site's colour. Decisions are edited in state\finding-decisions.json on the collector. Equipment that is out of support is on Lifecycle.

Backups

philadelphia: no backup for 2 days – the FortiGate did not answer (site unreachable). See Change → Backups.