Is mongodb application layer caching worth it?
I am using mongodb as a container for a collection of html help documents which are served by a REST server (cherry on top) using pymongo.
Since the database is relatively small (~ 50 documents) and rarely changes (only when someone edits my webserver interface, which runs on the same machine as the REST server), I wonder if I can speed it up with document caching ... Three options:
a) I could either cache locally (in a Python process) and then listen for, say, an invalidation trigger sent by my web server interface (for example, via a UNIX socket or signal or any other means).
b) I could cache via memcached and my webserver interface is invalid by deleting or updating said documents when something changes.
c) I couldn't do anything and relied on caching through the working set being automatically cached in memory by mongodb (is that?).
What do you think is the best strategy for such cases?
source to share
No one has answered this question yet
Check out similar questions: