Troubleshooting

Rule is pending but not deployed

Symptom: A rule was created or modified but has no effect on traffic.

Cause: Changes to rules are staged as Pending Changes and do not take effect on gateways until explicitly deployed.

Fix: Go to the Rules page and click Deploy pending changes. Check that the rule moves from the Pending Changes section to the Deployed Rules section after deploying.


Rule is deployed but the user is still blocked

Symptom: A firewall rule permits a user to access a resource, the rule is deployed, but the user still cannot reach it.

Before investigating the rule, determine whether the user can connect to the VPN location at all:

  • If the user cannot establish a VPN connection to the location, the problem is not the firewall rule. Access to a VPN location is controlled by the location's Allowed Groups setting, which is independent of firewall rules. A user who is not in an allowed group has no WireGuard configuration for that location and cannot connect regardless of any rules. Check the location's Allowed Groups configuration and ensure the user is a member of one of the permitted groups.

  • If the user can connect to VPN but still cannot reach the specific resource, the rule may not be correctly configured. Verify that:

    • The rule's destination matches the resource's IP address and the ports and protocols the user is trying to use.

    • The user, or a group they belong to, appears in the rule's Permissions.

    • The user is not being removed from the permitted set by a Restriction.

    • The rule is assigned to the correct location.


Rule is deployed but not yet active on the gateway

Symptom: A rule is in the Deployed Rules section in Core but is not reflected in the gateway's firewall.

Cause: Core pushes firewall configuration to gateways over the management connection. If that connection was unavailable at the time of deployment, the update was not delivered. Defguard guarantees eventual consistency - the configuration will be pushed the next time the connection is established.

Diagnosis: If your Gateway has the optional health endpoint configured, check its status:

  • 200 OK - Gateway is connected to Core. If the rule is still not active, redeploy the pending changes.

  • 503 Service Unavailable - Gateway is not connected to Core. Investigate the network path between Core and the Gateway's management port (TCP 50066). Once connectivity is restored, Core will automatically push the current firewall configuration on reconnect.


Disabling a rule has no immediate effect

Symptom: A rule was disabled in the UI but traffic it was controlling is still being permitted or blocked on the gateway.

Cause: Disabling a deployed rule creates a pending change, just like any other edit. The rule continues to be enforced on the gateway until the pending change is deployed.

Fix: Go to the Rules page, confirm the disabled rule appears in Pending Changes, and click Deploy pending changes.


Rule looks correct in the UI but produces unexpected firewall behaviour

Symptom: A rule is configured and deployed correctly, but traffic is permitted or blocked in ways that don't match what the UI shows.

Cause: When generating firewall rules, Defguard processes and merges all destination addresses across all rules before sending them to the gateway. This can produce rules with different IP notation than what was configured - for example, an IP range entered in the UI may appear as one or more subnets in the actual firewall rules. The logical coverage is identical, but the representation differs.

Additionally, if multiple rules cover overlapping destinations, their combined firewall rules may interact in ways that are not immediately obvious from the UI.

Diagnosis: Inspect the firewall rules applied on the gateway directly. See Implementation Details for the exact commands to list active rules for NFTables and Packet Filter, and for an explanation of how destination addresses are merged and represented.

Last updated

Was this helpful?