Maybe I lack conceptual understanding - but the diagram shows each cell with its own server, load balancer, and db. So I’m wondering what mechanisms are used to keep all those db’s in sync if clients are making requests that could be routed to any of n potential cells.
There could be cell replicas for high availability. But a customer gets routed to the same cell. And a cell could either represent a bounded context or the entire application.
If concurrency between users on different cells arise, perhaps reserved quota can be used. I don't know the right answer to this question.
can you give a more specific example?
Maybe I lack conceptual understanding - but the diagram shows each cell with its own server, load balancer, and db. So I’m wondering what mechanisms are used to keep all those db’s in sync if clients are making requests that could be routed to any of n potential cells.
There could be cell replicas for high availability. But a customer gets routed to the same cell. And a cell could either represent a bounded context or the entire application.
If concurrency between users on different cells arise, perhaps reserved quota can be used. I don't know the right answer to this question.