Standalone Machine Agent Settings for Docker Visibility
You can configure the following Docker Visibility settings on the Standalone Machine Agent. See also Standalone Machine Agent Configuration Property Reference.
The Machine Agent can monitor up to 150 running containers (depending on the pool size configuration, described below) and does not prioritize. The agent collects metrics from containers that run with a configurable regex. By default, this regex matches processes that are are running with the
Default = "
You can configure this option in the following file: <machine-agent-home>/extensions/DockerMonitoring/conf/DockerMonitoring.yml
Before you edit this file, please review "Editing YML Files: Important Notes" in Machine Agent Settings for Server Visibility.
Docker Metric Collector Pool Size
You can now configure Docker Visibility to monitor up to 150 containers at the same time. Previously, the supported maximum was 15 containers.
(New in 4.3.3) The Standalone Machine Agent can collect Memory and CPU metrics for Docker containers using cgroup data. This method is much faster than using the Docker API (the default metric-collection method). You can enable cgroup metric collection on all platforms except Amazon Linux OS.
This mode is disabled by default. To enable cgroup metric collection, run the agent with the following command-line option set to true:
java -Dappdynamics.docker.container.collection.cgroup.enabled=true -jar machineagent.jar
Controller Settings for Docker Visibility
For information about how to configure these settings, see Controller Settings for Standalone Machine Agents.
|Enable the Docker Monitoring feature.||true|
|Global limit for the number of containers to monitor per machine. You can specify this in the Administration Console as a Controller setting (all accounts) or as an Account setting for individual accounts. 120 is the maximum limit you can specify.||15|
Reuse SIM Machine entities to handle ephemeral environment. Support is limited to docker container machines for now.
If set to false, every new container will be considered a new machine.
|Mark the APM nodes as historical when a container running an app agent is stopped.||true|