Run this container on each of your Docker hosts. It will automatically create an agent instance inside Dataloop for each container on your host with basic operating system metrics under it.

To run this container:

API_KEY=<<Dataloop API Key>>
sudo docker run -d -e API_KEY=${API_KEY} \
--name=dataloop-docker \
--volume=/:/rootfs:ro \
--volume=/var/run:/var/run:rw \
--volume=/sys:/sys:ro \
--volume=/var/lib/docker/:/var/lib/docker:ro \
dataloop/dataloop-docker:latest

If you are behind a proxy server, you can pass the standard proxy environment variables to the docker container to have data passed through:

docker run -e HTTP_PROXY=http://proxy:port....

If you want to modify the container then feel free to submit a pull request. Below is the spec for what each script does.

A set of independent foreground processes that log to standard out that can be run under a supervisor.

All state is stored in Dataloop so these scripts can be run in ephemeral containers with no local storage.

• discover.py

Polls CAdvisor API and Dataloop. Ensures containers match agents via register and deregister API's.

• tag.py

Tags agents in Dataloop with their Docker Tags by matching container ID to agent name.

• metrics.py

Sends CAdvisor metrics to Dataloop via the Graphite endpoint every 10 seconds by matching container ID to agent name.

• check.py

Polls :8080/health_check for a 200 response code over docker internal network address and emits 0,1,2,3 to .health_check

docker run --rm -t -i -e API_KEY=$API_KEY \
--volume=/:/rootfs:ro \
--volume=/var/run:/var/run:rw \
--volume=/sys:/sys:ro \
--volume=/var/lib/docker/:/var/lib/docker:ro \
dataloop/dataloop-docker:latest /sbin/my_init -- bash -l