wavefront

Wavefront Output Plugin

This plugin writes to a Wavefront proxy, in Wavefront data format over TCP.

Configuration:

  ## Url for Wavefront Direct Ingestion or using HTTP with Wavefront Proxy
  ## If using Wavefront Proxy, also specify port. example: http://proxyserver:2878
  url = "https://metrics.wavefront.com"

  ## Authentication Token for Wavefront. Only required if using Direct Ingestion
  #token = "DUMMY_TOKEN"  
  
  ## DNS name of the wavefront proxy server. Do not use if url is specified
  #host = "wavefront.example.com"

  ## Port that the Wavefront proxy server listens on. Do not use if url is specified
  #port = 2878

  ## prefix for metrics keys
  #prefix = "my.specific.prefix."

  ## whether to use "value" for name of simple fields. default is false
  #simple_fields = false

  ## character to use between metric and field name.  default is . (dot)
  #metric_separator = "."

  ## Convert metric name paths to use metricSeparator character
  ## When true will convert all _ (underscore) characters in final metric name. default is true
  #convert_paths = true

  ## Use Strict rules to sanitize metric and tag names from invalid characters
  ## When enabled forward slash (/) and comma (,) will be accepted
  #use_strict = false
  
  ## Use Regex to sanitize metric and tag names from invalid characters
  ## Regex is more thorough, but significantly slower. default is false
  #use_regex = false

  ## point tags to use as the source name for Wavefront (if none found, host will be used)
  #source_override = ["hostname", "address", "agent_host", "node_host"]

  ## whether to convert boolean values to numeric values, with false -> 0.0 and true -> 1.0. default is true
  #convert_bool = true

  ## Truncate metric tags to a total of 254 characters for the tag name value. Wavefront will reject any 
  ## data point exceeding this limit if not truncated. Defaults to 'false' to provide backwards compatibility.
  #truncate_tags = false

  ## Flush the internal buffers after each batch. This effectively bypasses the background sending of metrics
  ## normally done by the Wavefront SDK. This can be used if you are experiencing buffer overruns. The sending 
  ## of metrics will block for a longer time, but this will be handled gracefully by the internal buffering in
  ## Telegraf.
  #immediate_flush = true

Convert Path & Metric Separator

If the convert_path option is true any _ in metric and field names will be converted to the metric_separator value. By default, to ease metrics browsing in the Wavefront UI, the convert_path option is true, and metric_separator is . (dot). Default integrations within Wavefront expect these values to be set to their defaults, however if converting from another platform it may be desirable to change these defaults.

Use Regex

Most illegal characters in the metric name are automatically converted to -.
The use_regex setting can be used to ensure all illegal characters are properly handled, but can lead to performance degradation.

Source Override

Often when collecting metrics from another system, you want to use the target system as the source, not the one running Telegraf. Many Telegraf plugins will identify the target source with a tag. The tag name can vary for different plugins. The source_override option will use the value specified in any of the listed tags if found. The tag names are checked in the same order as listed, and if found, the other tags will not be checked. If no tags specified are found, the default host tag will be used to identify the source of the metric.

Wavefront Data format

The expected input for Wavefront is specified in the following way:

<metric> <value> [<timestamp>] <source|host>=<sourceTagValue> [tagk1=tagv1 ...tagkN=tagvN]

More information about the Wavefront data format is available here

Allowed values for metrics

Wavefront allows integers and floats as input values. By default it also maps bool values to numeric, false -> 0.0, true -> 1.0. To map strings use the enum processor plugin.