For AI agents: A markdown version of this page is available at https://docs.datadoghq.com/tracing/trace_collection/runtime_config.md. A documentation index is available at /llms.txt.

Configuration at Runtime

This product is not supported for your selected Datadog site. ().

Overview

Configuration at runtime lets you modify Datadog SDK configuration from the Datadog UI, without needing to restart your application or service. You don’t need to wait for a new deployment or code change to update your configuration. Instead, update it right away with configuration at runtime.

Setup

Before you can use configuration at runtime, you must set up Remote Configuration. For more information, see Setting up Remote Configuration for Tracing.

Using configuration at runtime

The primary entry point for editing configuration at runtime is the SDK & Agent Configurations tab on the service page.

To make changes to a service’s configuration at runtime:

  1. Go to the Catalog in APM.
  2. Open the service’s Full Page.
  3. Click Service Info, and then go to Service Config > SDK & Agent Configurations.
  4. Select an environment.
  5. Click Edit on the configuration row you want to change. See supported configuration options for more details.
  6. Update the value in the edit modal and confirm.
The SDK Configurations table showing logs injection disabled across one instance.

In this example, you can see that Log Injection is disabled for the test environment across one instance. An instance refers to an instance of the Remote Configuration client. There should be one instance per process of your application.

You can tell when the configuration changes have been successfully applied by referencing the Updates applied to X instances text. In this example, the configuration applied successfully to one instance.

The Setup Guidance tab no longer includes an editing panel. It shows a banner that directs you to the SDK & Agent Configurations tab.

Supported configuration options

The following options are supported with configuration at runtime. The required tracer version is listed for each language:

OptionJavaNode.jsPython.NETRubyGoC++
Custom sampling rate
Set a global sampling rate for the library using DD_TRACE_SAMPLE_RATE.		1.17.0+4.11+ 3.32+ 2.45+2.4.0+2.33.0+1.13.0+1.59.0+0.2.0+
Log injection
Automatically inject trace correlation identifiers to correlate logs and traces by enabling DD_LOGS_INJECTION.		1.17.0+4.11+ 3.32+ 2.45+2.6.0+2.33.0+1.13.0+
HTTP header tags
Add HTTP header values as tags on traces using DD_TRACE_HEADER_TAGS.		1.17.0+4.11+ 3.32+ 2.45+2.6.0+2.33.0+1.13.0+1.59.0+
Custom span tags
Add specified tags to each span using DD_TAGS.		1.31.0+4.23.0+ 3.44.0+2.5.0+2.44.0+1.59.0+0.2.0+
Resource-based sampling
Set sampling rates by service and resource name, from the Datadog UI using DD_TRACE_SAMPLING_RULES.		1.34.0+5.16.0+2.9.0+2.53.2+2.0.0+1.64.0+0.2.2+
Adaptive Sampling
let Datadog control sampling rates on your behalf to match a configured monthly ingested volume budget.		1.34.0+5.16.0+2.9.6+2.54.0+2.0.0+1.68.0+0.2.2+

Further reading

Additional helpful documentation, links, and articles:

Remote ConfigurationDOCUMENTATION more more