Telegraf is able to serialize metrics into the following output data formats:

1. InfluxDB Line Protocol
2. JSON
3. Graphite

Telegraf metrics, like InfluxDB points, are a combination of four basic parts:

1. Measurement Name
2. Tags
3. Fields
4. Timestamp

In InfluxDB line protocol, these 4 parts are easily defined in textual form:

measurement_name[,tag1=val1,...]  field1=val1[,field2=val2,...]  [timestamp]

For Telegraf outputs that write textual data (such as kafka, mqtt, and file), InfluxDB line protocol was originally the only available output format. But now we are normalizing telegraf metric “serializers” into a plugin-like interface across all output plugins that can support it. You will be able to identify a plugin that supports different data formats by the presence of a data_format config option, for example, in the file output plugin:

[[outputs.file]]
  ## Files to write to, "stdout" is a specially handled file.
  files = ["stdout"]

  ## Data format to output.
  ## Each data format has its own unique set of configuration options, read
  ## more about them here:
  ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md
  data_format = "influx"

  ## Additional configuration options go here

Each data_format has an additional set of configuration options available, which I’ll go over below.

Influx:

There are no additional configuration options for InfluxDB line-protocol. The metrics are serialized directly into InfluxDB line-protocol.

Influx Configuration:

[[outputs.file]]
  ## Files to write to, "stdout" is a specially handled file.
  files = ["stdout", "/tmp/metrics.out"]

  ## Data format to output.
  ## Each data format has its own unique set of configuration options, read
  ## more about them here:
  ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md
  data_format = "influx"

Graphite:

The Graphite data format translates Telegraf metrics into dot buckets. A template can be specified for the output of Telegraf metrics into Graphite buckets. The default template is:

template = "host.tags.measurement.field"

In the above template, we have four parts:

1. host is a tag key. This can be any tag key that is in the Telegraf metric(s). If the key doesn’t exist, it will be ignored. If it does exist, the tag value will be filled in.
2. tags is a special keyword that outputs all remaining tag values, separated by dots and in alphabetical order (by tag key). These will be filled after all tag keys are filled.
3. measurement is a special keyword that outputs the measurement name.
4. field is a special keyword that outputs the field name.

Which means the following influx metric -> graphite conversion would happen:

cpu,cpu=cpu-total,dc=us-east-1,host=tars usage_idle=98.09,usage_user=0.89 1455320660004257758
=>
tars.cpu-total.us-east-1.cpu.usage_user 0.89 1455320690
tars.cpu-total.us-east-1.cpu.usage_idle 98.09 1455320690

Graphite Configuration:

[[outputs.file]]
  ## Files to write to, "stdout" is a specially handled file.
  files = ["stdout", "/tmp/metrics.out"]

  ## Data format to output.
  ## Each data format has its own unique set of configuration options, read
  ## more about them here:
  ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md
  data_format = "graphite"

  # prefix each graphite bucket
  prefix = "telegraf"
  # graphite template
  template = "host.tags.measurement.field"

JSON:

The JSON data format serialized Telegraf metrics in json format. The format is:

{
   "fields":{
      "field_1":30,
      "field_2":4,
      "field_N":59,
      "n_images":660
   },
   "name":"docker",
   "tags":{
      "host":"raynor"
   },
   "timestamp":1458229140
}

JSON Configuration:

[[outputs.file]]
  ## Files to write to, "stdout" is a specially handled file.
  files = ["stdout", "/tmp/metrics.out"]

  ## Data format to output.
  ## Each data format has its own unique set of configuration options, read
  ## more about them here:
  ## https://github.com/influxdata/telegraf/blob/master/docs/DATA_FORMATS_OUTPUT.md
  data_format = "json"
  json_timestamp_units = "1ns"

By default, the timestamp that is output in JSON data format serialized Telegraf metrics is in seconds. The precision of this timestamp can be adjusted for any output by adding the optional json_timestamp_units parameter to the configuration for that output. This parameter can be used to set the timestamp units to nanoseconds (ns), microseconds (us or µs), milliseconds (ms), or seconds (s). Note that this parameter will be truncated to the nearest power of 10 that, so if the json_timestamp_units are set to 15ms the timestamps for the JSON format serialized Telegraf metrics will be output in hundredths of a second (10ms).