Awesome Open Source
Awesome Open Source

Graphite Exporter

CircleCI Docker Repository on Quay Docker Pulls

An exporter for metrics exported in the Graphite plaintext protocol. It accepts data over both TCP and UDP, and transforms and exposes them for consumption by Prometheus.

This exporter is useful for exporting metrics from existing Graphite setups, as well as for metrics which are not covered by the core Prometheus exporters such as the Node Exporter.



Configure existing monitoring to send Graphite plaintext data to port 9109 on UDP or TCP. As a simple demonstration:

echo "test_tcp 1234 $(date +%s)" | nc localhost 9109
echo "test_udp 1234 $(date +%s)" | nc -u -w1 localhost 9109

Metrics will be available on http://localhost:9108/metrics.

To avoid using unbounded memory, metrics will be garbage collected five minutes after they are last pushed to. This is configurable with the --graphite.sample-expiry flag.

Graphite Tags

The graphite_exporter accepts metrics in the tagged carbon format. Labels specified in the mapping configuration take precedence over tags in the metric. In the case where there are valid and invalid tags supplied in one metric, the invalid tags will be dropped and the graphite_tag_parse_failures counter will be incremented. The exporter accepts inconsistent label sets, but this may cause issues querying the data in Prometheus.

Metric Mapping and Configuration

Please note there has been a breaking change in configuration after version 0.2.0. The YAML style config from statsd_exporter is now used. See conversion instructions below

YAML Config

The graphite_exporter can be configured to translate specific dot-separated graphite metrics into labeled Prometheus metrics via YAML configuration file. This file shares syntax and logic with statsd_exporter. Please follow the statsd_exporter documentation for usage information. However, graphite_exporter does not support all parsing features at this time. Any feature based on the 'timer_type' option will not function. Otherwise, regex matching, groups, match/drop behavior, should work as expected.

Metrics that don't match any mapping in the configuration file are translated into Prometheus metrics without any labels and with names in which every non-alphanumeric character except _ and : is replaced with _.

If you have a very large set of metrics you may want to skip the ones that don't match the mapping configuration. If that is the case you can force this behaviour using the --graphite.mapping-strict-match flag, and it will only store those metrics you really want.

An example mapping configuration:

- match: test.dispatcher.*.*.*
  name: dispatcher_events_total
    action: $2
    job: test_dispatcher
    outcome: $3
    processor: $1
- match: '*.signup.*.*'
  name: signup_events_total
    job: ${1}_server
    outcome: $3
    provider: $2
- match: 'servers\.(.*)\.networking\.subnetworks\.transmissions\.([a-z0-9-]+)\.(.*)'
  match_type: regex
  name: 'servers_networking_transmissions_${3}'
    hostname: ${1}
    device: ${2}

This would transform these example graphite metrics into Prometheus metrics as follows:

  => dispatcher_events_total{processor="FooProcessor", action="send", outcome="success", job="test_dispatcher"}

  => signup_events_total{provider="facebook", outcome="failure", job="foo_product_server"}
  => test_web__server_foo_bar{}

  => servers_networking_transmissions_failure_mean_rate{device="eth0",hostname="rack-003-server-c4de"}

Conversion from legacy configuration

If you have an existing config file using the legacy mapping syntax, you may use statsd-exporter-convert to update to the new YAML based syntax. Here we convert the old example synatx:

$ go get -u

$ cat example.conf


$ statsd-exporter-convert example.conf
- match: test.dispatcher.*.*.*
  name: dispatcher_events_total
    action: $2
    job: test_dispatcher
    outcome: $3
    processor: $1
- match: '*.signup.*.*'
  name: signup_events_total
    job: ${1}_server
    outcome: $3
    provider: $2

Using Docker

You can deploy this exporter using the prom/graphite-exporter Docker image.

For example:

docker pull prom/graphite-exporter

docker run -d -p 9108:9108 -p 9109:9109 -p 9109:9109/udp \
        -v ${PWD}/graphite_mapping.conf:/tmp/graphite_mapping.conf \
        prom/graphite-exporter --graphite.mapping-config=/tmp/graphite_mapping.conf

Experimental: Importing Whisper data

Import data from Graphite using the bundled getool. See getool create-blocks --help for usage.

To import long-term data in a reasonable amount of resources, increase the duration per generated TSDB block. The --block-duration must be a power of two in hours, e.g. 4h, 8h, and so on.

To merge the data into an existing Prometheus storage directory, start Prometheus with the --storage.tsdb.allow-overlapping-blocks flag.

Incompatibility with Graphite bridge

This exporter does not work in combination with the Java client or Python client Graphite bridge. In the transition to the Graphite data model and back, information is lost. Additionally, default metrics conflict between the client libraries and the exporter.

Instead, configure Prometheus to scrape your application directly, without the exporter in the middle. For batch or ephemeral jobs, use the pushgateway integration. If you absolutely must push, consider PushProx or the Grafana agent instead.

TLS and basic authentication

Graphite Exporter supports TLS and basic authentication. This enables better control of the various HTTP endpoints.

To use TLS and/or basic authentication, you need to pass a configuration file using the --web.config.file parameter. The format of the file is described in the exporter-toolkit repository.

Alternatives To Graphite_exporter
Select To Compare

Alternative Project Comparisons
Related Awesome Lists
Top Programming Languages
Top Projects

Get A Weekly Email With Trending Projects For These Topics
No Spam. Unsubscribe easily at any time.
Golang (158,575
Metrics (13,648
Monitoring (11,805
Yaml (10,305
Mapping (7,890
Prometheus (5,982
Graphite (1,656
Statsd (1,227
Observability (683