Troubleshoot a running infrastructure agent

You can troubleshoot our infrastructure agent using our newrelic-infra-ctl utility. This binary is also included in the installation package, in the same directory as the newrelic-infra default binary.

Upon receiving the newrelic-infra-ctl command, the agent:

  • Enables verbose logs for a period of five minutes, then reverts the log level to its previous setting.
  • Logs all the agent config options.
  • Logs all the integrations config options.
  • Executes a health check for every loaded integration. (A health check is an immediate execution of the integration with extra logs and output validation.)

Linux newrelic-infra-ctl

In Linux systems, the troubleshooting binary is /usr/bin/newrelic-infra-ctl, available in both the package manager or the tarball assisted install methods.

When running on Linux, newrelic-infra-ctl must be executed by either the root user or the same user running the newrelic-infra process.

The newrelic-infra-ctl binary can automatically detect the agent process running in the host. It can also detect whether the agent is running inside a container.

To change the default settings of newrelic-infra-ctl:

To change... Execute
pid newrelic-infra-ctl -pid 14580
cid (when using a containerized version of the agent) newrelic-infra-ctl -cid 8fddbcbb101c
docker-api-version newrelic-infra-ctl -docker-api-version 1.24

Windows newrelic-infra-ctl

In Windows, using the MSI installer, the troubleshooting binary is C:\Program Files\New Relic\newrelic-infra\newrelic-infra-ctl.exe.

In Windows, the agent process is always automatically detected. It does not depend on the pid or the cid.

For more help

If you need more help, check out these support and learning resources: