I have the same table as yours with potential to grow over 50 billion of
records once operational. But our hardware is currently very limited (8GB
RAM).
I concur with Tom Lane about the fact that partial indexes aren't really an
option, but what about partitioning?
I read from the Postgres docs that "The exact point at which a table will
benefit from partitioning depends on the application, although a rule of
thumb is that the size of the table should exceed the physical memory of the
database server."
http://www.postgresql.org/docs/current/static/ddl-partitioning.html
Now, a table with 500M records would exceed our RAM, so I wonder what impact
a table of 50G would have on simple lookup performance (i.e. source = fixed,
timestamp = range), taking into account that a global index would exceed our
RAM on some 1G records.
Did anyone do some testing? Is partitioning a viable option in such
scenario?
"Adrian von Bidder" <avbidder@fortytwo.ch> wrote in message
news:201003020849.19133@fortytwo.ch...