GoDoxy/next-release.md

GoDoxy v0.8 changes:

Breaking changes

• Removed redirect_to_https in config.yml, superseded by redirectHTTP as an entrypoint middleware

• New notification config format, support webhook notification, support multiple notification providers

  old

  providers:
    notification:
      gotify:
        url: ...
        token: ...
  

  new

  providers:
    notification:
      - name: gotify
        provider: gotify
        url: ...
        token: ...
      - name: discord
        provider: webhook
        url: https://discord.com/api/webhooks/...
        template: discord
  

  Webhook notification fields:

  Field	Description	Required	Allowed values
  name	name of the provider	Yes
  provider		Yes	webhook
  url	webhook URL	Yes	Full URL
  template	webhook template	No	empty, discord
  token	webhook token	No
  payload	webhook payload	No (if template is used)	valid json
  method	webhook request method	No	GET POST PUT
  mime_type	mime type	No
  color_mode	color mode	No	hex dec

  Available payload variables:

  Variable	Description	Format
  $title	message title	json string
  $message	message in markdown format	json string
  $fields	extra fields in json format	json object
  $color	embed color by color_mode	0xff0000 (hex) or 16711680 (dec)

Non-breaking changes

• services health notification now in markdown format like Uptime Kuma for both webhook and Gotify

• docker services now use docker container health status if possible, fallback to GoDoxy health check on failure / no docker health check, e.g.

  # docker compose
  services:
    app:
      ...
      container_name: app
      healthcheck:
        test: ["CMD-SHELL", "curl --fail http://localhost:8080 || exit 1"]
        interval: 5s
  

  Health check result will be equivalent to docker inspect --format='{{json .State.Health}}' app

• proxy.<alias>.path_patterns fully support http.ServeMux patterns [METHOD ][HOST]/[PATH] (See https://pkg.go.dev/net/http#hdr-Patterns-ServeMux)

• caching ACME private key in order to reuse ACME account, to prevent from ACME rate limit

• New: fully support string as inline YAML for docker labels

  services:
    app:
      ...
      # add '|' after colon ':' to treat it as string
      proxy.app: |
        scheme: http
        host: 10.0.0.254
        port: 80
        path_patterns:
          - GET /
          - POST /auth
        healthcheck:
          disabled: false
          path: /
          interval: 5s
      proxy.app1.healthcheck: |
        path: /ping
        use_get: true
      proxy.app1.load_balance: |
        link: app
        mode: ip_hash
  

• New: support entrypoint middlewares (applied to all routes, before route middlewares)

  entrypoint:
    middlewares:
      - use: CIDRWhitelist
        allow:
          - "127.0.0.1"
          - "10.0.0.0/8"
          - "192.168.0.0/16"
        status: 403
        message: "Forbidden"
  

• New: support exact host matching, i.e.

  # include file
  app1.domain.tld:
    host: 10.0.0.1
  
  # docker compose
  services:
    app1:
      ...
      proxy.aliases: app1.domain.tld
  

  will only match exactly app1.domain.tld match_domains in config will be ignored for this route

• New: support host matching without a subdomain, i.e.

  app1:
    host: 10.0.0.1
  

  will now also match app1.tld

• New: support x-properties (will be ignored, like in docker compose), useful with YAML anchor e.g.

  x-proxy: &proxy # this will be ignored in GoDoxy
    scheme: https
    healthcheck:
      disable: true
    middlewares:
      hideXForwarded:
      modifyRequest:
        setHeaders:
          Host: $req_host
  
  api.openai.com:
    <<: *proxy # extends from x-proxy
    host: api.openai.com
  api.groq.com:
    <<: *proxy # extends from x-proxy
    host: api.groq.com
  

• new middleware name aliases:

  • modifyRequest = request
  • modifyResponse = response

• New: support $ variables in request and response middlewares (like nginx config)

  • $req_method: request http method
  • $req_scheme: request URL scheme (http/https)
  • $req_host: request host without port
  • $req_port: request port
  • $req_addr: request host with port (if present)
  • $req_path: request URL path
  • $req_query: raw query string
  • $req_url: full request URL
  • $req_uri: request URI (encoded path?query)
  • $req_content_type: request Content-Type header
  • $req_content_length: length of request body (if present)
  • $remote_addr: client's remote address (may changed by middlewares like RealIP and CloudflareRealIP)
  • $remote_host: client's remote ip parse from $remote_addr
  • $remote_port: client's remote port parse from $remote_addr (may be empty)
  • $resp_content_type: response Content-Type header
  • $resp_content_length: length response body
  • $status_code: response status code
  • $upstream_name: upstream server name (alias)
  • $upstream_scheme: upstream server scheme
  • $upstream_host: upstream server host
  • $upstream_port: upstream server port
  • $upstream_addr: upstream server address with port (if present)
  • $upstream_url: full upstream server URL
  • $header(name): get request header by name
  • $resp_header(name): get response header by name
  • $arg(name): get URL query parameter by name

• New: Access Logging (entrypoint and per route), i.e.

  # config.yml
  entrypoint:
    access_log:
      format: json # common, combined, json
      path: /app/logs/access.json.log
      filters:
        cidr:
          negative: true # no log for local requests
          values:
            - 127.0.0.1/32
            - 172.0.0.0/8
            - 192.168.0.0/16
            - 10.0.0.0/16
      fields:
        headers:
          default: drop # drop app headers in log
          config: # keep only these
            X-Real-Ip: keep
            CF-Connecting-Ip: keep
            X-Forwarded-For: keep
  
  # include file
  # same as above but under route config
  app:
    access_log:
      format: json # common, combined, json
      ...
  
  # docker labels
  labels:
    proxy.app.access_log: |
      format: json
      path: /app/logs/access.json.log
      filters:
        cidr:
          negative: true
          values:
            - 127.0.0.1/32
            - 172.0.0.0/8
            - 192.168.0.0/16
            - 10.0.0.0/16
  

  mount logs directory before setting this

Fixes

• duplicated notification after config reload
• timeout was defaulted to 0 in some cases causing health check to fail
• redirectHTTP middleware may not work on non standard http port
• various other small bugs
• realIP and cloudflareRealIP middlewares
• prometheus metrics gone after a single route reload
• upgraded dependencies to the latest