Multi tenant usecase of atlas

I need to provide collection for each user, but on atlas creating too many collection on single cluster can slow down writes so i will just spin new cluster programmatically after hitting 300 collections. The problem will be managing too many connections on the backend. To what extent I can scale this strategy?

Hi @Arnav_Singh1 and welcome in the MongoDB Community :muscle: !

Here is a topic that will help a bit I think:


I am creating a mini database. so, each user will have different index and schema. so, the suggested strategy won’t be useful. What will be the pitfalls of my strategy of spinning new cluster after reaching 300 to 400 collection