Agent Runtime Configuration Management

If you are running Agent 6.19+/7.19+, you can dynamically change some settings at runtime without having to restart the Agent to account for the configuration change.

Note: Changes made dynamically do not persist. They are lost as soon as the Agent is restarted.

Use the command config list-runtime to list configuration parameters that can be changed at runtime. See the table below for the complete command on different platforms.

PlatformCommand
Dockerdocker exec datadog-agent agent config list-runtime
macOSdatadog-agent config list-runtime
CentOSsudo datadog-agent config list-runtime
Debiansudo datadog-agent config list-runtime
Kuberneteskubectl exec <POD_NAME> agent config list-runtime
Fedorasudo datadog-agent config list-runtime
Redhatsudo datadog-agent config list-runtime
Susesudo datadog-agent config list-runtime
Sourcesudo datadog-agent config list-runtime
WindowsConsult the dedicated Windows documentation

One parameter that can be changed at runtime is the log level. It is convenient for debug purposes in a containerized environment, where the Agent configuration cannot be changed without having to destroy then recreate the container running the Agent. To dynamically set the log level to debug on a Kubernetes deployment, invoke the following command:

kubectl exec <POD_NAME> agent config set log_level debug

It is possible to get the current value of runtime-editable settings by using config get <SETTING>. For example, to get the current log level on a Linux system, use:

sudo datadog-agent config get log_level

The complete runtime configuration can also be show by using the config command.

Further Reading

Additional helpful documentation, links, and articles:

PREVIEWING: may/unit-testing