Action Items

Explore Redis as in-memory data store to efficiently manage tasks
Basic: Implement a global click counter for each user that rate limits them after 10 clicks in 10 seconds
- Rate Limit: Use a queue in Redis to store the request counts for each user
- Click Counting: Implement a task queue to handle click count as dispatched tasks
Advanced: Create a collaborative whiteboard that multiple users can draw on

Retrospective

Implementing the click counter was fairly easy
Getting my head around how Redis works and how to implement stuff on it wasn’t hard but took some time
The project asked us to use lists/queues but this is dumb and using a counter makes way more sense. I first implemented it with counter and it was trivial then I tried implementing it with lists and I could do it but it was very inefficient so I switched back to counters.
Registering unique users and keeping track of individual user clicks in reddis took a lot of time. principally because I was trying to both register a user AND process their clicks in the same API call. I was trying to think of elaborate ways in which I could register the user variable and then also use it and was constantly running into the problem of redefining the user each time a click was made… it took me way too long to realise that I should seperate these two functions. So now a useeffect hook runs at the start of page load which calls a register-user API which create a unique user and stores them in redis, and then onclick another api call does the actual click counting for that user. much more elegant solution.
Deploying a backend on railway is pretty easy. But connecting it to reddis is a bit tricky if you don’t know about shared variables. Deploying a frontend on railway SUCKs
Deploying a frontend on vercel is really easy.

Blockers

Was feeling tired and a little burnt out from lots of late nights - need some solid rest over the weekend
Spent almost 2 hours trying to deploy the frontend on railway… gave up and switched to vercel - deployed in <5mins :)

Learning

working with different aspects of redis: basic set/get, counters, lists and queueing, rate limiting
deploying a redis db on the cloud and connecting backend to it

Wins

Got the project deployed - backend and redis on railway + vite frontend on vercel
Getting faster at deployments now

PRs