Currently, If there is use-case that the data needs to brought from tables resting in different databases of a same cluster. We have to create n connections for n databases.
But being in same cluster, one should be able to access different databases with a single connection otherwise the connection list gets long and messy.