PostgreSQL maintains a collection of multiple databases (database cluster) in a single server process. Immediately after installation, it contains two template databases, template0 and template1, and a regular database called postgres.

When you think of a cluster, you might think of a state that does things like mirroring between separate entities, but a PostgreSQL database cluster is just a collection of databases, without the ability to mirror between individual databases. For high availability, use techniques such as replication to replicate the entire database cluster.

You can set a different character set for each database. When you create a new database cluster with the initdb command, you specify the character set to use for template1, and postgres, which is created by copying it, will have the same character set. Later, when you add a database, for example with the createdb command, it defaults to the same character set as template1, but you can set any character set by specifying template0 as the template database.

User information is created in the global area of the database cluster and is common to all databases in the cluster. Settings, such as whether each user is allowed to connect and whether operations on individual objects are allowed, can be set on each database. However, CREATE USER is a user creation for a database cluster, not an operation on individual database.

Database user information is not included in each database and cannot be backed up by the pg_dump command. Although the pg_dumpall command backs up all databases in the cluster, you can specify the -g option to back up only the information contained in the global region of the cluster (in particular, role and tablespace information).

So the correct answers are A, C, and E.