Simon Riggs wrote:
>
> As Ron says, until we have partitioning, it isn't as useful for DW.
Well, it's a bit stronger than what I said. The last big DW I dealt
with did have previous quarter's data archived into different tables,
and those could be marked read-only. Also, quite a few of our
ID<->Value lookup tables have a fixed or slowly changing set of
values (products, states, etc); and those would benefit from index scans.
But yes, it'd be an even nicer feature with partitioning, since then
you could hypothetically keep inserting into a mostly-read-only table
and mark partitions read-only as needed.
I think I was trying to say it'd be even more useful for DW with
partitioning - and I hope that when partitioning is being designed
it considers the possibility of taking advantage of read-only
partitions if we happen to implement read-only tables.
