versipellis

Versipellis Configuration Reference

[sender.http] Sub-Section

url - server address, in the form "http[s]://host[:port][/path][?query]"

• Required
• If the host is an IPv6 address, it must be enclosed in square brackets, e.g., [fe80::1]
• Query parameters may be specified here as a suffix (e.g., ?param1=value1&param2=value2), but it’s recommended to specify them in query (see the notes there)

method - HTTP request method, a.k.a. verb, for sending data

• Optional
• Default: "POST"
• Options (case insensitive): "PATCH", "POST", "PUT"

query - HTTP query parameters

• Optional
• Query parameters may be specified as a suffix in url, but it’s recommended to specify them here:
  • Values in query override values in url, if they have the same parameter names
  • Values in query are allowed to contain dynamic expressions which are evaluated during runtime (although this is not implemented yet), while url must be static

• The TOML file format supports multiple representation options for key-value pairs:

  [sender.http]
  # ...
  query = { param1 = "value1", param2 = "value2" }
  

  [sender.http]
  # ...
  query = {
      param1 = "value1",
      param2 = "value2",
  }
  

  [sender.http]
  # ...
  query.param1 = "value1"
  query.param2 = "value2"
  

  [sender.http]
  # ...
  
  [sender.http.query]
  param1 = "value1"
  param2 = "value2"
  

headers - HTTP request headers

• Optional

• The TOML file format does not allow multiple instances of the same field, so if you want to specify a header with multiple values, specify them as a single comma-separated string:

  accept = "text/plain, application/json, application/xml"
  

• The TOML file format supports multiple representation options for key-value pairs:

  [sender.http]
  # ...
  headers = { header1 = "value1", header2 = "value2" }
  

  [sender.http]
  # ...
  headers = {
      header1 = "value1",
      header2 = "value2",
  }
  

  [sender.http]
  # ...
  headers.header1 = "value1"
  headers.header2 = "value2"
  

  [sender.http]
  # ...
  
  [sender.http.headers]
  header1 = "value1"
  header2 = "value2"
  

timeout - maximum amount of time for each HTTP client request to complete

• Optional
• Default: "5s" (5 seconds)
• Format: string containing decimal numbers, each with a unit suffix, e.g., m (minutes), and s (seconds)
• Special case: "0" and negative values (e.g., "-1s") = no client-side timeout

format - convert structured data into a specific content type

• Optional
• Default: the source’s format setting, if it has one, otherwise "ndjson"
• Options:
  • "json" (an array if the data contains multiple objects, otherwise an object)
  • "json_array" (always an array, even when there are fewer than 2 objects)
  • "ndjson" (NDJSON, a.k.a. JSONL)
• This overrides the format setting of the source collector or receiver
• This also sets the HTTP header Content-Type automatically, unless you override it in headers
• Exception: opaque byte payloads and proxied HTTP requests and responses are sent as-is, ignoring this setting

concurrency_limit - how many requests are allowed to be in progress at any time

• Optional
• Default: 100
• Valid range: any positive integer between 1 (no concurrency) and 200
• Integer values out of bounds in either direction are normalized to the maximum
• If the sender is at its concurrency limit when a new request needs to start,
  the request is delayed until a slot is available, not rejected/dropped
• Each request keeps its slot until it completes, including its retries

[!WARNING]

• Concurrency limit (at any time, unrelated to duration or retries) ≠ requests per second
• Data that is still blocked when a shutdown begins is dropped, instead of delaying the shutdown
  (data that doesn’t exceed the concurrency limit, including the last pending batch, is still sent)

[sender.http.tls] Sub-Section

• Optional for HTTP/1.1, required for HTTP/2
• See this page

[sender.http.retries] Sub-Section

• Optional
• See this page

[sender.http.batch] Sub-Section

• Optional
• Default: disabled, immediate dispatch without size limits
• See this page

This site is open source. Improve this page.