http

HTTP Processor Plugin

The http processor plugin performs HTTP requests for each event. This plugin requires a parser and a serializer.

Unlike the HTTP output, this plugin uses a configured label as the request path (it is optional), not the event routing key.

Please note that in multiline configuration, the HTTP client is shared between processors in the set.

Configuration

[[processors]]
  [processors.http]
    # target host, required
    host = "http://localhost:9100"

    # list of fallback hosts
    # if all `retry_attempts` to perform request to `host` failed, fallbacks will be used
    fallbacks = [ "http://fallback.local:9100", "http://anotherfallback.local:9100" ]

    # request method, required
    method = "POST"

    # time limit for requests made by client
    # zero means no limit
    timeout = "10s"

    # maximum amount of time an idle (keep-alive) connection will remain idle before closing itself
    # zero means no limit
    idle_conn_timeout = "1m"

    # maximum number of idle (keep-alive) connections across all hosts
    # zero means no limit
    max_idle_conns = 10

    # result codes, means request performed successfully
    success_codes = [ 200, 201, 204 ]

    # regular expression to match response body
    # if configured, response code must be in `success_codes` OR body must match this regexp
    success_body = ".*success but with strange code.*"

    # label, which value will be used as a request path, if configured
    path_label = "request_uri"

    # label, which value will be used as a request method, if configured
    # if not, or event has no label, `method` will be used
    method_label = "request_method"

    # field, which content will be used as a request body after serialization, if configured
    request_body_from = "path.to.request.field"

    # field, that will be added to event with response body after parsing, if configured
    response_body_to = "path.to.response.body"

    # field, that will be added to event with response code, if configured
    response_code_to = "path.to.response.code"

    # maximum number of attempts to perform request
    # before events will be marked as failed
    # 
    # only requests that ended with `success_codes`
    # will NOT be retried
    retry_attempts = 0 # zero for endless attempts

    # interval between retries to perform bulk request
    retry_after = "5s"

    ## TLS configuration
    # if true, TLS client will be used
    tls_enable = false
    # trusted root certificates for server
    tls_ca_file = "/etc/neptunus/ca.pem"
    # used for TLS client certificate authentication
    tls_key_file = "/etc/neptunus/key.pem"
    tls_cert_file = "/etc/neptunus/cert.pem"
    # minimum TLS version, not limited by default
    tls_min_version = "TLS12"
    # send the specified TLS server name via SNI
    tls_server_name = "exmple.svc.local"
    # use TLS but skip chain & host verification
    tls_insecure_skip_verify = false

    # static headers that will be used on each request
    [processors.http.headers]
      authorization = "@{envs:BEARER_TOKEN}"

    # a "header <- label" map
    # if event label exists, it will be added as a request header
    # if "headers" already has same one, it will be overwritten
    [processors.http.headerlabels]
      custom_header = "my_label_name"

    # a "label <- header" map
    # if response header exists, it will be saved as configured label
    [processors.http.labelheaders]
      my_label_name = "x-custom-header"

    # a "param <- field" map
    # if event field exists, it will be added as a request urlparam
    # target field must not be a map, but slices allowed, if slice not contains maps
    [processors.http.paramfields]
      custom_param = "my.field.path"

    [processors.http.serializer]
      type = "json"
      data_only = true

    [processors.http.parser]
      type = "json"
      split_array = true