Create an Integration-Recommended Monitor
Overview
Datadog Monitors track key metrics, so you can efficiently monitor your infrastructure and integrations. Datadog provides a set of out-of-the-box monitors for many features and integrations. View these monitors in your Monitors Template list.
Create an out-of-the-box monitor to help users find value in your Datadog integration. This guide provides steps for creating an integration-recommended monitor and best practices to follow during the creation process.
To create a Datadog integration, see Create a New Integration.
Steps to create a recommended monitor
Build a monitor JSON Schema
Navigate to Monitors > New Monitor and create a new monitor.
Follow the best practices listed in this guide to configure your monitor.
Click Export Monitor.
Check Select to export as a recommended monitor.
Click Copy to use the JSON schema of your configured monitor.
Save the copied schema to a JSON file and name it according to your monitor title. For example, your_integration_name_alert.json.
In the monitor JSON file, fill out the Title, Description, and Tags. For more information, see Configuration best practices.
Open a pull request
Save the monitor JSON file to your integration’s assets/monitors folder. Add the asset to your manifest.json file. See Integrations Assets Reference for more information about your integration’s file structure and manifest file.
Open a pull request (PR) to add your recommended monitor JSON file and updated manifest file to the corresponding integration folder either in the integrations-extras GitHub repository or Marketplace GitHub repository.
After it’s approved, Datadog merges the PR and your integration-recommended monitor is pushed to production.
Verify your monitor in production
To see the out-of-the-box monitor, the relevant integration tile must be Installed in Datadog.
Find your monitor in the Monitors Template list. Ensure logos render correctly on the Monitors Template lists page.
Configuration best practices
In addition to the monitor definition, the Title, Description, and Tags fields are required for recommended monitors. For more information, see the documentation on configuring a monitor.
| Description | Examples |
|---|
| Title | Allows users to quickly understand the underlying failure mode the alert is covering. The rule for the title is starting with an object and following with a verb. Do not use template variables. | - Pods are failing
- Cache usage is high
- Error rate is high
- Disk space is low |
| Description | Provides extra context around the failure mode and also about the impact this mode can have on the system. It should be concise and allow users to understand at a glance whether it is relevant or not for them to create a monitor out of it. | Title: Unacknowledged Messages are higher than usual
Description: Unacked messages are those that have been delivered to a consumer but have not been acknowledged as processed or handled. This monitor tracks the ratio of unacked messages to avoid potential bottlenecks which could lead to delays in message processing. |
| Tags | Set to “integration:<app_id>”. | See other available monitor tags here. |
Below is an example of a well-defined monitor:
Further reading
Additional helpful documentation, links, and articles:
