Rate Limiting Shadow

# Rate Limiting Shadow Mode
#
# Observe a rate limit before enforcing it. The limiter evaluates every
# request and counts the ones it would have rejected, but lets them all
# through, so a new limit can be tuned on live traffic without returning
# a single 429.
#
# Responses still carry the X-RateLimit-* headers, so a drained bucket
# shows up as "X-RateLimit-Remaining: 0" on a 200. Would-be rejections
# are counted in praxis_rate_limit_limited_total{shadow="true"} on the
# metrics listener and logged as shadow decisions. Remove
# `shadow: true` to start enforcing.
#
# Usage:
#   cargo run -p praxis-proxy -- -c examples/configs/traffic-management/rate-limiting-shadow.yaml
#   for i in $(seq 1 10); do curl -si http://localhost:8080/ | grep -i x-ratelimit-remaining; done
#   curl -s http://localhost:9901/metrics | grep rate_limit_limited

admin:
  metrics_address: "127.0.0.1:9901"

listeners:
  - name: web
    address: "127.0.0.1:8080"
    filter_chains:
      - shadow-limited

filter_chains:
  - name: shadow-limited
    filters:
      - filter: rate_limit
        mode: per_ip
        rate: 1              # demo value; tune against the metric
        burst: 2             # demo value; tune against the metric
        shadow: true         # observe only: count would-be rejections, never 429

      - filter: router
        routes:
          - path_prefix: "/"
            cluster: backend

      - filter: load_balancer
        clusters:
          - name: backend
            endpoints:
              - "127.0.0.1:3000"

insecure_options:
  allow_private_endpoints: true # example proxies to local backends