A common pattern for microservices is that each should reference an independent data store. To that end, I am planning on allocating a single mongoDB instance, but provisioning a separate database for each microservice (all databases, that is, co-located within the same mongoDB instance).
This partitions the data store by microservice, but also introduces potential contention issues (over network bandwidth, connection limits to mongoDB, etc).
Is there a best practice with respect to mongoDB deployment / configuration in this context?