A chat with a startup founder made me question heavy dependencies. Why not use plain JSON files with basic lookups and in-memory caching instead of setting up elaborate database systems?
hey, i used json files for a small proj which worked fine untill user growth made caching act weird. simple is ok at first, but be prepped for scale hiccups
Utilizing plain JSON files for storage can be an efficient solution in scenarios where the application’s scope is limited and data volume is minimal. In my experience working on small-scale projects, this approach allowed for rapid development and easier maintenance without the complexity of a full-fledged database. However, it is important to consider that as the system grows or demands concurrent access, file storage may introduce challenges related to data consistency and concurrency. Weighing these trade-offs is crucial before making the switch, especially when long-term scalability is a potential requirement.
hey, im all in for simple solutions but wonder if using json could struggle when many users access it simultanously. has anyone tried mixing file storage with a caching layer? how does that work out when load increases, or are there hidden issues?
Drawing from my experience, using JSON files with an in-memory cache can be ideal for rapid prototyping and projects with limited data. I implemented a similar setup in a device management tool where the ease of file-based storage expedited development and debugging. However, I observed that when user numbers increased, managing file locks and consistency became challenging even with caching in place. It is essential to plan for a graceful migration to a more robust solution if significant scaling or concurrent writes are expected in the future.