In the experiment section of the updated GrowthBook, how can we continue to test in the test environment after creating a new experiment and linking it to previous features? Before upgrading GrowthBook, we used "features" to create experiments, which included both production and test environments. When we needed to update the experiment configuration, we would first test it in the test environment. If there were no issues, we would then use the "copy to production" feature to officially deploy it. Due to the need to use "start new phase" to redistribute traffic, we upgraded GrowthBook. However, we found that after the upgrade, when we create a new experiment in the "experiment" section and link it to the previous features, we can no longer distinguish between the production and test environments. This means that redistributing traffic in the test environment also affects the production environment. I would like to know whether I haven't found the correct way to use the platform after the upgrade, or if the new version itself does not support distinguishing between different environments within "experiment." Being able to test changes in experiment configuration and traffic before making them is still very important to us.

I will have to check with the team about that

Experiments have no awareness of environment. Any feature that references them can apply feature-level environment gating in front of any experiment targeting (such as a rollout rule before hitting the experiment rule), but any experiment level targeting or traffic redistribution happens across all environments. What folks often do is clone their experiment for different environments ("Duplicate" is available from the experiment's top-right drop down menu).