You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Now that the api test are merged, we still need to improve some parts, so that testing can be streamlined:
Github actions are not working correctly. This is because API tests needs ray, and Github CI machines are not powerful enough to host our ray container (actually, we even disable ray with --disable_ray and RAY_ENABLED=False in the configs for testing). The solution is to host our own ray service on our GPU machines.
During development, it should be possible to run the tests locally without any disturbance. Right now, I need to reset my database(s) to have fresh environments for tests. Actually, I want all tests to be run on a seperate sql, elasticsearch, weaviate databases so that my development environment is NOT disturbed. Raffael once prepared a solution for this in run_tests.sh. and a corresponding env.testing.example file. We have to improve this solution, optimally so that everything is configured correctly in our launch.json. Not sure if we really need two new weaviate and ES containers, we just have to make sure that everything is cleaned, after the tests run. In conftest.py, there are already hooks for this (STARTUP_DONE:) and pyteset_sessionfinish().
The text was updated successfully, but these errors were encountered:
Now that the api test are merged, we still need to improve some parts, so that testing can be streamlined:
Github actions are not working correctly. This is because API tests needs ray, and Github CI machines are not powerful enough to host our ray container (actually, we even disable ray with --disable_ray and RAY_ENABLED=False in the configs for testing). The solution is to host our own ray service on our GPU machines.
During development, it should be possible to run the tests locally without any disturbance. Right now, I need to reset my database(s) to have fresh environments for tests. Actually, I want all tests to be run on a seperate sql, elasticsearch, weaviate databases so that my development environment is NOT disturbed. Raffael once prepared a solution for this in run_tests.sh. and a corresponding env.testing.example file. We have to improve this solution, optimally so that everything is configured correctly in our launch.json. Not sure if we really need two new weaviate and ES containers, we just have to make sure that everything is cleaned, after the tests run. In conftest.py, there are already hooks for this (STARTUP_DONE:) and pyteset_sessionfinish().
The text was updated successfully, but these errors were encountered: