Linkerd Proxy Admin API

The Linkerd proxy exposes an admin HTTP server on each meshed pod, providing health check endpoints, readiness probes, Prometheus-compatible metrics, and runtime diagnostic information. By default this server listens on port 4191.

OpenAPI Specification

linkerd-proxy-admin-openapi.yml Raw ↑
openapi: 3.1.0
info:
  title: Linkerd Proxy Admin API
  description: >-
    The Linkerd proxy exposes an admin HTTP server on each meshed pod,
    providing health check endpoints, readiness probes, Prometheus-compatible
    metrics, and runtime diagnostic information. By default this server
    listens on port 4191.
  version: 2.x
  contact:
    name: Linkerd
    url: https://linkerd.io/
  license:
    name: Apache 2.0
    url: https://www.apache.org/licenses/LICENSE-2.0
externalDocs:
  description: Linkerd Proxy Configuration Reference
  url: https://linkerd.io/2-edge/reference/proxy-configuration/
servers:
  - url: http://localhost:4191
    description: Linkerd proxy admin server (default port)
paths:
  /metrics:
    get:
      operationId: getProxyMetrics
      summary: Linkerd Get proxy metrics
      description: >-
        Returns Prometheus-compatible metrics for the Linkerd proxy,
        including request counts, latency histograms, TCP connection
        statistics, and TLS handshake information.
      tags:
        - Metrics
      responses:
        '200':
          description: Prometheus-formatted metrics
          content:
            text/plain:
              schema:
                type: string
                description: Prometheus exposition format metrics
  /ready:
    get:
      operationId: getProxyReadiness
      summary: Linkerd Proxy readiness check
      description: >-
        Returns the readiness status of the Linkerd proxy. Used by
        Kubernetes as a readiness probe to determine if the proxy
        is ready to accept traffic.
      tags:
        - Health
      responses:
        '200':
          description: Proxy is ready
          content:
            text/plain:
              schema:
                type: string
        '503':
          description: Proxy is not ready
  /live:
    get:
      operationId: getProxyLiveness
      summary: Linkerd Proxy liveness check
      description: >-
        Returns the liveness status of the Linkerd proxy. Used by
        Kubernetes as a liveness probe to determine if the proxy
        process is alive.
      tags:
        - Health
      responses:
        '200':
          description: Proxy is alive
          content:
            text/plain:
              schema:
                type: string
  /shutdown:
    post:
      operationId: shutdownProxy
      summary: Linkerd Initiate proxy shutdown
      description: >-
        Triggers a graceful shutdown of the Linkerd proxy. This endpoint
        is used during pod termination to ensure the proxy drains
        connections before exiting.
      tags:
        - Lifecycle
      responses:
        '200':
          description: Shutdown initiated
          content:
            text/plain:
              schema:
                type: string
tags:
  - name: Health
    description: Health and readiness check endpoints
  - name: Lifecycle
    description: Proxy lifecycle management endpoints
  - name: Metrics
    description: Prometheus metrics endpoints