phil campaigne wrote: > > Does this also mean that you cannot physically split a database over > multiple physical servers? > Phil Physically, yes, logically no. Suppose, for example, you have two servers A and B, both running PostgreSQL with corresponding databases dbA and dbB. Depending on what you are trying to accomplish, you can put table 1 in dbA and table2 in dbB. The application users don't know and don't care where the tables reside. If you want to hide this from the programmers as well, you can hide that with a very thin access layer. Finally, if you need to join the two tables, you can even do that with the dblink add-on. Finally, if you are using Java, ObjectWeb has a RAIDb driver that can transparently distribute databases over a cluster of servers. See http://c-jdbc.objectweb.org/. -- Guy Rouillier ---------------------------(end of broadcast)--------------------------- TIP 9: the planner will ignore your desire to choose an index scan if your joining column's datatypes do not match