Seeing hundreds of SSL VPN login failed attempts? Start with Fortinet's own settings, then block the IPs already attacking VPN portals with a threat feed in a local-in policy, so they never reach the login page.
Start a 14-day free trialOr test 200 entries first: free, no signupHow do I stop brute force on a FortiGate SSL VPN? Layer Fortinet's own controls with a block list: require MFA, tune login-attempt-limit and login-block-time, limit who can reach the portal by country or source, then deny known attacking IPs with an external threat feed in a local-in policy on the WAN interface. On FortiOS 7.6.3 and later, protect the IPsec gateway the same way.
Which blocklist works with FortiGate SSL VPN? Any plain-text IPv4 list that FortiOS can load as an IP Address threat feed. ThreatListPro is one VPN-focused option: 5.4M+ IPv4 addresses from real attacks on live production Palo Alto GlobalProtect VPN portals in the US, checked for new attackers every 6 hours. Use it as the source of a local-in deny policy when SSL VPN listens on the WAN interface (FortiOS 7.2.4 or later in the 7.2 branch, or 7.4 and later). For SSL VPN on a loopback interface reached through a VIP, use it in a firewall policy instead. FortiOS 7.0 can’t use a threat feed in a local-in policy, and a normal firewall policy does not filter traffic to a VPN bound directly to the WAN interface. Check trusted sources before you enforce it, and keep MFA and firmware updates.
Free test list (200 entries, no signup): https://raw.githubusercontent.com/pcchiu2023/threatlistpro-blocklist/main/data/sample/blocklist.txt · GitHub repository with setup examples
It is a very common question in Fortinet's own forums, common enough that Fortinet keeps a dedicated mitigation resource list for it: the VPN event log fills with failed SSL VPN logins, usually for usernames that don't exist, from IP addresses you have never seen. It is automated credential guessing, also called password spraying.
All of these are free and built into FortiOS. Each one helps; none of them stops the attempts arriving on its own. Keep FortiOS patched as well: attackers are exploiting a FortiOS CAPWAP flaw to plant a remote-access trojan, so check which FortiOS version fixes CVE-2025-25249.
| Control | What it does | Limit |
|---|---|---|
| MFA (FortiToken, or SAML with your identity provider) | A guessed password alone no longer gets in | Attempts still arrive and fill the logs |
login-attempt-limit / login-block-time (config vpn ssl settings) | Blocks a source IP for a while after repeated failures | Distributed attacks stay under it; Fortinet notes it doesn't apply to SAML user groups |
| Restrict by country or source (geography addresses, local-in policy) | Cuts most of the noise if your users are in known countries | Attackers using IPs inside your allowed countries still get through |
| SSL VPN realms | Sends logins only to the intended LDAP or RADIUS server, so random attempts don't lock out accounts | Reduces lockouts, not attempts |
| Automation stitch on failed logins | Adds an offending IP to a block group automatically | Reacts after that IP has already tried; Fortinet's FortiAnalyzer variant needs FortiAnalyzer |
| Disable web mode, move the port | Removes the web login page and some default-port scanning (from 7.6.3, web mode, now Agentless VPN, is the only SSL VPN mode left) | Determined scanners find the new port |
| Migrate to IPsec VPN | Fortinet's recommendation; required for full-tunnel remote access from 7.6.3 | The IPsec gateway still faces login attempts |
Every control above reacts to failures on your own FortiGate. A threat feed of IPs already seen attacking VPN portals blocks them before their first attempt on yours. Fortinet supports this directly: an IP threat feed can be the source address of a local-in policy, and a local-in deny drops the listed IP's packets before they reach the SSL VPN service, so the login page never loads for that IP.
ThreatListPro is one such feed: 5.4M+ IPv4 addresses, built from real attacks on live production Palo Alto GlobalProtect VPN portals in the US and checked for new attackers every 6 hours. It is plain text, one IPv4 address or CIDR per line, the format FortiOS expects for an IP Address threat feed.
FortiOS 7.2.4 or later (or 7.4.0 or later). FortiOS 7.0 can't use a threat feed in a local-in policy; there, use the feed in a firewall policy with SSL VPN on a loopback interface (see the note in step 4). Local-in policies are CLI-only before 7.6.0.
Test with the free 200-entry sample above, or use your ThreatListPro URL from the welcome email:https://api.threatlistpro.com/v1/YOUR_API_KEY/blocklist.txt
Go to Security Fabric > External Connectors > Create New > IP Address. Name it ThreatListPro, paste the URL and set the refresh rate to 60 minutes. In a Security Fabric, configure the feed on each FortiGate that needs it. Every GUI field and CLI option is explained in FortiGate threat feed setup.
config system external-resource
edit "ThreatListPro"
set type address
set resource "https://api.threatlistpro.com/v1/YOUR_API_KEY/blocklist.txt"
set refresh-rate 60
next
end
Before you enforce it, compare the source IPs of your successful SSL VPN logins over the last 30 days with the list. If a trusted network appears, allow it with a higher-priority local-in policy first.
Create an address for the WAN IP and a service for the SSL VPN port, then deny the feed. Replace wan1, the IP and the port with your own. Pick an ID not shown by show firewall local-in-policy; local-in policies match top-down, so put any allow-list policy above this one:
config firewall address
edit "sslvpn_wan"
set subnet 203.0.113.10 255.255.255.255
next
end
config firewall service custom
edit "sslvpn_port"
set tcp-portrange 443
set udp-portrange 443
next
end
config firewall local-in-policy
edit <unused-id>
set intf "wan1"
set srcaddr "ThreatListPro"
set dstaddr "sslvpn_wan"
set service "sslvpn_port"
set schedule "always"
set action deny
next
end
SSL VPN on a loopback interface? Then the traffic passes a normal firewall policy from WAN to the loopback (through a VIP). Use the feed there instead, either as the source of a deny policy or with set srcaddr-negate enable on the allow policy, as Fortinet's technical tip shows.
Moved to IPsec on FortiOS 7.6.3+? Use the same local-in policy with the IKE and ESP services (IKE is UDP 500 and 4500), plus the TCP port if you use IKE over TCP. Like any local-in block, it also blocks existing tunnels from listed IPs.
Run diagnose sys external-address-resource list ThreatListPro to see the loaded entries, or diagnose sys external-resource stats for every feed's valid and error line counts. Then watch Log & Report > System Events > VPN Events: failed logins from listed IPs should stop.
| Login attempt limit | Automation stitch | Country restriction | Threat feed (ThreatListPro) | |
|---|---|---|---|---|
| Blocks before the first attempt on your FortiGate | ✗ | ✗ | For blocked countries | For listed IPs |
| Handles low-rate attempts spread across many IPs | ✗ | Only after each IP fails | Outside allowed countries | For listed IPs |
| Cost | Free | Free (FortiAnalyzer variant needs FortiAnalyzer) | Free | $19.99/mo, or the free 200-entry sample |
They work best together: MFA and the login limit on the account side, country restriction and a threat feed at the edge.
An SSL VPN portal answers on the internet, so automated tools find it and try common usernames and leaked passwords around the clock. The attempts usually come from many IP addresses, each trying only a few times, which keeps them under per-IP lockout thresholds. They show up in Log & Report > System Events > VPN Events as failed SSL VPN logins.
Only partly. login-attempt-limit and login-block-time under config vpn ssl settings block a source IP after repeated failures, so they slow one attacker down. A distributed attack spreads its attempts across many IPs, so most of them never reach the limit, and Fortinet notes the setting does not apply to SAML user groups.
Yes. Fortinet documents a local-in policy with an external IP threat feed as the source address to block those IPs from reaching an SSL VPN that listens on the WAN interface. This needs FortiOS 7.2.4 or later (or 7.4.0 or later); FortiOS 7.0 doesn't support it; there, a firewall policy can use the feed only if SSL VPN listens on a loopback interface reached through a VIP. Before 7.6.0 local-in policies are configured from the CLI only; from 7.6.0 they can also be created in the GUI. If SSL VPN listens on a loopback interface, use a firewall policy instead.
Yes. From FortiOS 7.6.3, SSL VPN tunnel mode is removed on all FortiGate models (web mode remains as Agentless VPN on supported models), and Fortinet recommends migrating to IPsec VPN. The IPsec gateway is attacked the same way, and the same threat feed can be used in a local-in policy for the IKE and ESP services, plus the TCP port if you use IKE over TCP.
Every 60 minutes is plenty. ThreatListPro is checked for new attackers every 6 hours, so refreshing more often than hourly only adds requests. The list covers 5.4M+ IPv4 addresses in only a few thousand entries, well under the FortiGate threat feed entry limit.
Yes. A free 200-entry sample with no signup is published on GitHub at https://raw.githubusercontent.com/pcchiu2023/threatlistpro-blocklist/main/data/sample/blocklist.txt. The full list is $19.99/month with a 14-day free trial and no card needed.
One threat feed URL in a local-in policy. 14-day free trial, no card needed.
Start a 14-day free trialOr test 200 entries first: free, no signupThreatListPro provides a standard IP blocklist feed compatible with any firewall that supports external lists.
Learn more about IP blocklists, VPN security, and how ThreatListPro compares to alternatives.