Skip to main content

Common Scenarios

This page summarizes common ForgeDNS policy combinations.

1. Cache + Upstream Forwarding​

Use cache in front of forward for the default fast path:

- tag: seq_main
  type: sequence
  args:
    - exec: "$cache_main"
    - matches: "!$has_resp"
      exec: "$forward_main"

2. Domain-Based Split Routing​

Use domain_set plus qname-style matching to steer queries to different upstreams.

Typical uses:

  • domestic vs. global upstream split
  • ad / tracking filtering
  • service-specific routing

3. LAN Policy + ECS​

Use client_ip plus ecs_handler when answers should depend on client network identity or geolocation hints.

4. Fallback Chains​

Use fallback when:

  • one upstream is low-latency but incomplete
  • one upstream is trusted but slower
  • you need a controlled degraded mode

5. Local Synthetic Answers​

Use:

  • hosts
  • arbitrary
  • redirect

These are useful for internal names, sinkholes, testing, or synthetic records.

6. DNS-Driven System Integration​

Use executors such as:

  • ipset
  • nftset
  • mikrotik
  • reverse_lookup

These allow DNS answers to drive routing, firewall, and inventory state outside the DNS server itself.

7. Observability​

Use:

  • query_summary
  • metrics_collector
  • plugin-specific HTTP endpoints

Keep observability available, but avoid pushing every side effect directly into the latency-critical path unless necessary.