Debugging containers or runner.py itself
The first step in debugging a measurement workflow is to turn the
flag of the
When you call the
runner.py locally it will turn into a steppable mode where
you contine to the next step by pressing enter.
You can then enter one of the containers to see if the required services are running correctly.
An example call would be:
docker exec -it MY_CONTAINER_NAME bash
Some container do not have
sh, which has less capabilities,
should be available in most cases.
Debbuging containers via HTTP / exposed ports
If entering the container looks fine and you need to access them through some of their
exposed ports (ex. via Browser through HTTP) turn on the
--allow-unsafe flag to bind
the ports specified in the
Debugging metric providers
To see if the Metric Providers → are working correctly you have two options:
- Start them manually from their respective folder under
/tools/metric-providers/.../static-binaryand look if the output is as expected
- Turn on the
--no-file-cleanupswitch to see if the files generated in
/tmp/green-metrics-tool/[...].logare in expected format