For me this topic is also from interest.
Would be great if @sunnyson could share insights on this topic.
For example one DB containing “massive”-data but is mostly only used to read from with another DB containing strapi + some other “not so massive”-data.