Caddy

Monitor Caddy web server — request rates, response latencies, active connections, and TLS handshake stats — using Caddy's built-in Prometheus metrics endpoint.

Pattern: Caddy /metrics → Prometheus scrape → xScaler remote_write

---

Prerequisites

• Caddy 2.x
• metrics directive enabled in the Caddyfile or JSON config
• Admin API accessible (default localhost:2019)

---

Enabling Caddy metrics

Add the metrics directive to your Caddyfile global block:

{
  admin localhost:2019
  metrics
}

example.com {
  reverse_proxy localhost:8080
}

Caddy exposes metrics at http://localhost:2019/metrics.

---

Option A — Prometheus Exporter

Scrape the Caddy admin API metrics endpoint directly:

scrape_configs:
  - job_name: caddy
    static_configs:
      - targets: ["localhost:2019"]
    metrics_path: /metrics

remote_write:
  - url: https://euw1-01.m.xscalerlabs.com/api/v1/push
    authorization:
      credentials: <token>
    headers:
      X-Scope-OrgID: <tenant-id>

---

Option B — Grafana Alloy

prometheus.scrape "caddy" {
  targets      = [{ __address__ = "localhost:2019" }]
  metrics_path = "/metrics"
  forward_to   = [prometheus.remote_write.xscaler.receiver]
}

prometheus.remote_write "xscaler" {
  endpoint {
    url = "https://euw1-01.m.xscalerlabs.com/api/v1/push"
    authorization {
      type        = "Bearer"
      credentials = "<token>"
    }
    headers = { "X-Scope-OrgID" = "<tenant-id>" }
  }
}

---

Option C — OpenTelemetry Collector

receivers:
  prometheus:
    config:
      scrape_configs:
        - job_name: caddy
          static_configs:
            - targets: ["localhost:2019"]
          metrics_path: /metrics

processors:
  memory_limiter:
    check_interval: 1s
    limit_mib: 256
  batch:
    timeout: 10s

exporters:
  otlphttp/xscaler:
    endpoint: https://euw1-01.m.xscalerlabs.com
    headers:
      Authorization: "Bearer <token>"
      X-Scope-OrgID: "<tenant-id>"
    compression: gzip

service:
  pipelines:
    metrics:
      receivers:  [prometheus]
      processors: [memory_limiter, batch]
      exporters:  [otlphttp/xscaler]

---

Logs

Collect Caddy access and error events via systemd journal. Add the following to your Alloy config:

loki.source.journal "caddy_journal" {
  forward_to    = [loki.write.xscaler.receiver]
  relabel_rules = loki.relabel.caddy_journal.rules
  labels = {
    job      = "integrations/caddy",
    instance = constants.hostname,
  }
}

loki.relabel "caddy_journal" {
  forward_to = []
  rule {
    source_labels = ["__journal__systemd_unit"]
    target_label  = "unit"
  }
}

loki.write "xscaler" {
  endpoint {
    url = "https://euw1-01.l.xscalerlabs.com/api/v1/logs/push"

    http_client_config {
      authorization {
        type        = "Bearer"
        credentials = env("XSCALER_TOKEN")
      }
    }

    headers = { "X-Scope-OrgID" = env("XSCALER_TENANT_ID") }
  }
}

Key metrics

Metric	Description
caddy_http_requests_in_flight	Current number of active HTTP requests being handled
caddy_http_request_duration_seconds	Histogram of HTTP request processing time
caddy_http_response_size_bytes	Histogram of HTTP response body size
caddy_tls_handshake_duration_seconds	Histogram of TLS handshake completion time
caddy_reverse_proxy_upstreams_healthy_total	Number of healthy upstream backends