Instability around fixtures that add environment variables to MCG's pods

[sagihirshfeld]

add_env_vars_to_noobaa_core_fixture and add_env_vars_to_noobaa_endpoint_fixture are patching a STS/Deployment, then resets and waits for the relevant pod to restart.

There are two issues with this:

1. Patching the STS/Deployment triggers the restart of the pods by itself, so the following manual restart is redundant.
2. The health check is implicit in the method that resets the pods, and sometimes incomplete. This caused multiple race conditions already in our pipeline.