I have the following scenario:
- I have multiple instances of a Flask application behind a load balancer serving an API for a front end web client. This flask application interfaces with a database to query and insert data.
- Under certain conditions, an event in my system will cause all the clients to make the exact same query to the API which then results in many calls to a database.
- In order to prevent unnecessary DB queries, I would like to implement a cache using Redis as the back end storage solution. I'm Using the Flask-Caching library with the RedisCache backend. All instances of the API service share the same Redis backend.
My question is this: Will the Redis backend block processes trying to write and read from the cache? Each request to the cached endpoint can take up to 3 seconds (worst case scenario). While waiting for a response, many additional requests may be made to that endpoint. Will they end up waiting for the cache to be written to by the original request?
I know Redis has a way to create locks, and I would assume that a cache would implement some of lock around cached values, but I cannot find any documentation as to how Flask-Cache interfaces with Redis.
