Hello all,I'm researching about more advanced Node.js servers and would like some input from the community on scaling very large applications.From what I can tell, since Node.js is single-threaded you must scale horizontally, via server clusters (and sharded databases) for large amounts of data. So this brings up a lot of questions.At what point do I need to scale across regions? If I have a robust server located in NYC, how slow will requests be in Mumbai, for example? For an app that requires real-time interaction between users with Websockets, will request latency make the app inoperable? Is keeping CPU% down really all that matters to keep server requests fast, even across the world? (I understand you want to distribute servers for redundancy and fault tolerance).What do the code deploys look like across various regions of the world? Should I just use Kubernettes and/or Docker and/or Jenkins for automating these? What are the best tools for this?What are the best tools for health checking throughout these various regions?Are there any other critical pieces of this puzzle i'm missing?Anyone here have experience with this kind of scaling, or architecture? It would be great to get expert advice on this. :)
Submitted May 07, 2020 at 03:14PM by _autoencoder_
No comments:
Post a Comment