-
Notifications
You must be signed in to change notification settings - Fork 9
Description
As opposed to dogpile.cache, where there are examples of invalidating the cache for particular queries, how is Cache invalidated on this project?
-
Is it done automatically? EDIT: (I saw there's an invalidate method, but when is it invoked?)
-
If so, should I be worried about using raw sql (bulk inserts, bulk updates)? how to handle those cases?
EDIT:
3a) is the query caching useful from request to request? it seems that the namespace is created per query. namespace contains a random number...
3b) What's the ideal scenario for the cache? How can I manually ensure I keep the cache live for the most time possible while also making sure that's invalidated when the data is not up-to-date (assuming my app is the only one that modifies the Database.
Thanks in advance.
EDIT 2:
- doing db.session.commit() will make an allegedly "cached" query to be called again. Even though the commit was for a totally different entity than the one involved in the cached query. Am I missing something? Can't I just get a cache and invalidate it manually?