Time to hit F5 on some OS concepts

A lot stuffs happened during this break. I'll scratch em down here later cz it isn't time for it.

It's Sunday 2:30 am and I'm not gonna lie, I couldn't sleep in peace for the past 2 days as my mind has been haunted by an issue.
One of my hobby app is firing a lot of requests to the Google Cloud Datastore. Technically speaking, my app has been well optimized for least requests - and it is doing it right. It turns out the issue is normal: the more people uses the app the more hits on the Datastore and each hit costs.
Even though Google offers an economic awesome pricing scheme on Datastore, I've to account for higher hit ratio. I'm not rich, I've to fix this.

I'm thinking of some kind of a caching mechanism. I've found some npm modules out their, unfortunately they don't have the right balls. So decided to give a try to make a pluggable solution that works just like how the OS manages hdd-ram-cache stack on a computer. But, my app is not built to run on any computer, but on server. They are different. It's necessary to consider the consistency of data over multiple instance deployments and parallel processors, and how this is gonna behave over a load balancer.

So it's time to press F5 on my OS and Parallel processing classes and jump right into code.
I'll be publishing it once it is ready.

Peace ✌️️

Popular posts from this blog

Talky Messenger Documentation & Setup

How to pass variables to res.render() in Node.js

4 tiny questions you fail easily.