I am currently prototyping a geo distributed microservice backend (Restful API running on GKE). In this regard, I would kindly like to evaluate CockroachDB.
I am new to CockroachDB and wanted to know as best/common practice should I have:
CockroachDB in the same node as my service(web server), so differents pods living on same node.
I should have a dedicated Cockroach DB cluster isolates from services(web server) clusters.
None of above
My question might come from missing some fundamental concept arpund how CockroachDB remains consistent accross geo distributed architecture. ( how communication is achieve accross locations on GKE)