I have this https://github.com/splitio/python-client#getting-started exact same code in my Flask endpoint and it works on dev where it is a single process environment.
But on QA, we are using uwsgi to spawn new threads and deal with more requests. So according to this https://help.split.io/hc/en-us/articles/360020359652#sdk-architecture , I was assuming that if we want good performance and not hit the API all the time then we should use caching layer in the middle.
But looks like that is not the case. As right now, When we restart the uwsgi and make the 1st request it gets all the flags and its states as they are in Split.io dashboard. Any new Feature Flag that are added or any rules changes don't reflect. I think the right word is they don't get synchronized. Is that they case or did I miss something?
To start with we want some basic setup and were avoiding cache as that might give us false results.
Please sign in to leave a comment.