It's a good idea to partition large distributed enterprises into autonomous subsystems.
The segregation of write and read database schemas, and indeed write and read databases can help to avoid one subsystem becoming unavailable because another became unavailable.
This is based on the important observation that data owned by one service (meaning it's responsible for the write schema and the write database) tends to only be read by other services.
Furthermore often only a small subset is read. This involves a logical view.
Data replication makes the system more robust to outages of individual components.