should we use redis?

[brvn0]

Since each frontend client is asking the backend like 16x per second for all categories (and each request to the backend results in a db query!!) I thought about caching responses of common queries
I mean the problem could also be resolved by just saving data and not requesting it 16 times per second but we could use redis which makes our project cooler :)

(redis is an in memory key value store and therefore much faster and more efficient than postgres)

heres a log where one client asks the backend for categories and about.json exactly every 60ms

[2025/07/03 11:31:59] [DEBUG] request completed latency=0.030588 method=GET path=/v1/about.json status=200
[2025/07/03 11:31:59] [DEBUG] request completed latency=0.480522 method=GET path=/v1/categories/ status=200
[2025/07/03 11:32:05] [DEBUG] Getting category list
[2025/07/03 11:32:05] [DEBUG] request completed latency=0.027672 method=GET path=/v1/about.json status=200
[2025/07/03 11:32:05] [DEBUG] request completed latency=0.419468 method=GET path=/v1/categories/ status=200
[2025/07/03 11:32:11] [DEBUG] Getting category list
[2025/07/03 11:32:11] [DEBUG] request completed latency=0.025448 method=GET path=/v1/about.json status=200
[2025/07/03 11:32:11] [DEBUG] request completed latency=0.444023 method=GET path=/v1/categories/ status=200
[2025/07/03 11:32:17] [DEBUG] Getting category list
[2025/07/03 11:32:17] [DEBUG] request completed latency=0.027692 method=GET path=/v1/about.json status=200
[2025/07/03 11:32:17] [DEBUG] request completed latency=0.4048 method=GET path=/v1/categories/ status=200
[2025/07/03 11:32:23] [DEBUG] Getting category list
[2025/07/03 11:32:23] [DEBUG] request completed latency=0.02687 method=GET path=/v1/about.json status=200
[2025/07/03 11:32:23] [DEBUG] request completed latency=0.375044 method=GET path=/v1/categories/ status=200

[brvn0]

@Gregor-Gottschewski @Julessh05 what do you think?

[Gregor-Gottschewski]

How easy is it to set up?

[brvn0]

pretty easy, i would do it, we can use it on a per-request basis (not caching recent advertisements but caching user attributes or individual advertisement attributes)
cache invalidation is also possible so removing stale data should not be a problem

[juliandevelops]

Multiple data fetching queries most likely happens due to rerenders on the client side.
This could be solved using a third party query and caching library such as tanstack query in the frontend. This could be a feature, but is not a priority to the frontend.
Backend caching could be a nice solution to this

[juliandevelops]

Since each frontend client is asking the backend like 16x per second for all categories (and each request to the backend results in a db query!!) I thought about caching responses of common queries I mean the problem could also be resolved by just saving data and not requesting it 16 times per second but we could use redis which makes our project cooler :)

(redis is an in memory key value store and therefore much faster and more efficient than postgres)

heres a log where one client asks the backend for categories and about.json exactly every 60ms

[2025/07/03 11:31:59] [DEBUG] request completed latency=0.030588 method=GET path=/v1/about.json status=200
[2025/07/03 11:31:59] [DEBUG] request completed latency=0.480522 method=GET path=/v1/categories/ status=200
[2025/07/03 11:32:05] [DEBUG] Getting category list
[2025/07/03 11:32:05] [DEBUG] request completed latency=0.027672 method=GET path=/v1/about.json status=200
[2025/07/03 11:32:05] [DEBUG] request completed latency=0.419468 method=GET path=/v1/categories/ status=200
[2025/07/03 11:32:11] [DEBUG] Getting category list
[2025/07/03 11:32:11] [DEBUG] request completed latency=0.025448 method=GET path=/v1/about.json status=200
[2025/07/03 11:32:11] [DEBUG] request completed latency=0.444023 method=GET path=/v1/categories/ status=200
[2025/07/03 11:32:17] [DEBUG] Getting category list
[2025/07/03 11:32:17] [DEBUG] request completed latency=0.027692 method=GET path=/v1/about.json status=200
[2025/07/03 11:32:17] [DEBUG] request completed latency=0.4048 method=GET path=/v1/categories/ status=200
[2025/07/03 11:32:23] [DEBUG] Getting category list
[2025/07/03 11:32:23] [DEBUG] request completed latency=0.02687 method=GET path=/v1/about.json status=200
[2025/07/03 11:32:23] [DEBUG] request completed latency=0.375044 method=GET path=/v1/categories/ status=200

I cannot replicate this behaviour, also, it's 6 seconds, isn't it?
On my terminal, an initial loading is completed, but after that, no further activities unless navigating to new page

[brvn0]

Well you are right lmao its every 6 seconds, I'll not implement redis xD