Skip to main content

Datadog/OpenCost Plugin Notes


Support for the Datadog plugin is available in the current 1.110.0 release.

These instructions are very rough and will continue to get revised as we approach release.

Plugin Architecture:

Plugin Downloading

The OpenCost Helm chart uses an init container to install the plugins the user has configured. The init container downloads the plugins and puts them in a local emptydir that is mounted in the standard OpenCost container. The plugins themselves run in the existing OpenCost process.

OpenCost uses this approach of dynamically downloading the plugins to avoid container bloat. A single plugin could double the base OpenCost container image size - this allows OpenCost to scale to support many plugins by only downloading and installing the ones the user selects. This does not increase the number of running containers in an OpenCost installation, it only adds an init container.

OpenCost Datadog Plugin Setup Guide

  1. Obtain plugin binary
    1. Visit and choose the most recent release
    2. Download the correct binary for your architecture/OS
    3. Create a sibling directory to OpenCost called plugins
    4. In plugins directory, create two sub directories: bin and config
    5. Put the plugin binary in the bin dir, ensure it is executable
  2. Create datadog_config.json in the config dir, with the following contents:
"datadog_site": "",
"datadog_api_key": "c508d4fd3d126abbbbdc2fe96b0f6613",
"datadog_app_key": "f357b1f4efefb0870109e0d1aa0cb437b5f10ab9"
  1. Ensure you set the following env vars:
    1. PLUGIN_EXECUTABLE_DIR Should have the full path to the bin dir you set up
    2. PLUGIN_CONFIG_DIR Should have the full path to the config dir
    3. CUSTOM_COST_ENABLED Set to “true”
    4. Also, set LOG_LEVEL to ‘debug’
  2. Add in any other env vars you normally pass to OpenCost
  3. Start the app
  4. You should see a lot of debug output from the Datadog plugin
  5. After the ingestion completes, the info should be available in the repo objects, for both hourly and daily

Testing the Helm chart

  1. Support has been merged into the OpenCost Helm chart
  2. Update your local OpenCost Helm values file with the following settings:
loglevel: debug
enabled: true
enabled: true
fullImageName: curlimages/curl:latest
folder: /opt/opencost/plugin
# leave this commented to always download most recent version of plugins
# leave this commented to always download most recent version of plugins
# the list of enabled plugins
enabledPlugins: []
# - datadog
# pre-existing secret for plugin configuration
configSecret: ""

datadog: |
"datadog_site": "",
"datadog_api_key": "c508d4fd3d126abbbbdc2fe96b0f6613",
"datadog_app_key": "f357b1f4efefb0870109e0d1aa0cb437b5f10ab9"

cloudProviderApiKey: "AIzaSyDXQPG_MHUEy9neR7stolq6l0ujXmjJlvk"
# -- Exporter container image name
repository: guestbook-227502/opencost
# -- Exporter container image tag
# @default -- `""` (use appVersion in Chart.yaml)
tag: plugin-spike-03062024
  1. Confirm that looking at the pod logs shows Datadog queries going through
Documentation Distributed under CC BY 4.0.  The Linux Foundation® (TLF) has registered trademarks and uses trademarks. For a list of TLF trademarks, see: Trademark Usage.