SQL is highly unpopular, might just agree 😵😵 le me : Writing 3NF forms for every entity and assuming interviewier is fine with me using SQL Databases only to be asked why not slap Cassandra on this and call it a day. 🤦♂ Happened a while back. Was waiting on this podcast. API versioning is something I will definitely be taking away. I myself have never talked about that or though of it to be addressed in an interview. 😃
Thanks Rajat for watching the video and sharing your experience. Yeah - API versioning is critical and there are multiple ways for doing that. I suggest taking a look at API Design Masterclass where we talk about different ways of versioning an API.
Great video, got amazing insights. One thing I wanted to dig deeper is choosing compute in capacity planning. Are there some standard methods to determine if my system with "x" cores and "y" GB RAM is good enough to serve my application to "n" number of users. Load testing is the first thing which comes to my mind, just curious to know how such decision are taken in general.
Hi Vishwas, Thanks for watching the video and showing your interest. Specifically for your question, I would say typically we figure out what type of workload it is i.e. IO, Memory or CPU intensive. Once we know that we try to make some calculations based on the req/s for IO, data size for memory and algorithm complexity for CPU intensive tasks. Then best is to test the workload on a single machine and figure out the breaking point. Once you know the breaking point it is typically easily to figure out bottlenecks and also find the scaling model for your application. Hope it helps, Good luck 🙏🏻😊
