Open ribalba opened 10 months ago
Is this still current, given your yesterday discovery that we start containers and use the reporters to derive data?
I see atm no way to parallelize this properly ... at least if you also want proper data output from the metric providers ...
There are also loads of tests that don't require the containers. All the schema/ parameter checking for example. We could have two categories of test and such speed up the testing. Really nothing we need right now but the tests running so long is quite annoying.
I will schedule it in for next week with Dan
Improvement done here: https://github.com/green-coding-berlin/green-metrics-tool/pull/637
Test time is now roughly 3 mins.
The onlything that actually takes up time is if the containers are started.
it would be technically possible to start the testing container with a different Id and thus effectively parallelize the tests.
Leaving this issue open for further inspection if tests balloon up again.
Currently the tests run for quite some time and there are some tests we could run in parallel. There is the great
pytest-xdist
package that uses workers to run tests in parallel. At some stage we should maybe look through all the tests and categorize them into two groups so we can seep up the ones that we can parallelize.Very low prio though.