Class FleetAutomationApi

Cancel an active deployment and stop all pending operations. When you cancel a deployment:

• All pending operations on hosts that haven't started yet are stopped.
• Operations currently in progress on hosts may complete or be interrupted, depending on their current status.
• Configuration changes or package upgrades already applied to hosts are not rolled back.

After cancellation, you can view the final state of the deployment using the GET endpoint to see which hosts were successfully updated before the cancellation.

Only deployments with a pending or running status can be canceled. Returns a 400 if the deployment is not in a cancelable status. Returns a 404 if no deployment matches the specified ID or if you do not have access to it.

Create a new deployment to apply configuration changes to a fleet of hosts matching the specified filter query.

This endpoint supports two types of configuration operations:

• merge-patch: Merges the provided patch data with the existing configuration file, creating the file if it doesn't exist.
• delete: Removes the specified configuration file from the target hosts.

You can optionally use target_packages to apply the configuration change only to specific package versions.

The deployment is created and started automatically. You can specify multiple configuration operations to execute in order on each target host. Use the filter query to target specific hosts using the Datadog query syntax.

Set dry_run to true to validate the configuration and resolve target hosts and packages without deploying anything. A dry run returns a 200 with the validation result instead of creating and starting a deployment.

Returns a 400 if filter_query or config_operations is missing, a target package is missing a name or version or cannot be resolved, the configuration fails validation, or the filter query does not match any host eligible for the deployment.

Create and immediately start a new package upgrade on hosts matching the specified filter query.

This endpoint allows you to upgrade the Datadog Agent to a specific version on hosts matching the specified filter query.

The deployment is created and started automatically. The system:

1. Identifies all hosts matching the filter query.
2. Validates that the specified version is available.
3. Begins rolling out the package upgrade to the target hosts.

Returns a 400 if filter_query or target_packages is missing, a target package is missing a name or version, or the filter query does not match any host eligible for the upgrade. Returns a 409 if a conflicting upgrade is already running on one or more target hosts.

Create a new schedule for automated package upgrades.

Schedules define when and how often to automatically deploy package upgrades to a fleet of hosts. Each schedule includes:

• A filter query to select target hosts
• A recurrence rule defining maintenance windows
• A version strategy (e.g., always latest, or N versions behind latest)

When the schedule triggers during a maintenance window, it automatically creates a deployment that upgrades the Datadog Agent to the specified version on all matching hosts.

Delete a schedule permanently.

When you delete a schedule:

• The schedule is permanently removed and will no longer create deployments
• Any deployments already created by this schedule are not affected
• This action cannot be undone

If you want to temporarily stop a schedule from creating deployments, consider updating its status to "inactive" instead of deleting it.

Retrieve detailed information about a specific Datadog Agent.

By default, only agent_infos is returned. Use the include query parameter to request additional data: integrations and/or configuration_files.

Retrieve detailed information about a specific deployment, including its current status, configuration operations, and per-host execution status.

Returns a 404 if no deployment matches the given ID or if you do not have access to it.

Retrieve detailed information about a specific schedule by its unique identifier.

Retrieve a paginated list of tracers for a specific agent.

This endpoint returns tracers associated with a given agent key, identified by the agent's hostname. Use this to discover telemetry-derived service names for a particular host.

Retrieve the list of Datadog Agent versions available for deployment.

Returns 200 with an empty data array if the Agent package exists in the catalog but has no available versions, and 404 only if the Agent package itself is absent from the catalog.

Retrieve a paginated list of Datadog Agents.

Returns agents with support for pagination, sorting, and filtering. Use page_number and page_size to navigate pages, filter to narrow by field values, and tags to filter by agent tags.

Retrieve a paginated list of all deployments for fleet automation.

Retrieve all upgrade schedules for the organization.

Schedules automate package upgrades by defining maintenance windows and recurrence rules. Each schedule automatically creates deployments based on its configuration.

Retrieve a paginated list of all fleet tracers.

This endpoint returns telemetry-derived service names from the SDK telemetry pipeline. These names may differ from span-derived names in APM and are useful for querying service library configurations. Use the page_number and page_size query parameters to paginate through results.

Manually trigger a schedule to immediately create and start a deployment.

This endpoint allows you to manually initiate a deployment using the schedule's configuration, without waiting for the next scheduled maintenance window. This is useful for:

• Testing a schedule before it runs automatically
• Performing an emergency update outside the regular maintenance window
• Creating an ad-hoc deployment with the same settings as a schedule

The deployment is created immediately with:

• The same filter query as the schedule
• The package version determined by the schedule's version strategy
• All matching hosts as targets

The manually triggered deployment is independent of the schedule and does not affect the schedule's normal recurrence pattern.

Partially update a schedule by providing only the fields you want to change.

This endpoint allows you to modify specific attributes of a schedule without affecting other fields. Common use cases include:

• Changing the schedule status between active and inactive
• Updating the maintenance window times
• Modifying the filter query to target different hosts
• Adjusting the version strategy

Only include the fields you want to update in the request body. All fields are optional in a PATCH request.