On Tue, May 20, 2003 at 12:58:41AM -0400, Tom Lane wrote:
> "Jim C. Nasby" <jim@nasby.net> writes:
> > I'm having the same problem...
>
> How many rows would you expect this command to insert?
>
> You might be running into the fact that pending-trigger-call lists are
> kept in memory, but it's hard to tell without knowing the number of
> rows involved...
35M
I re-wrote the query to do a subselect instead of a full join (which
changed the access plan from a hash-join to a subquery), and it looks
like it might finish, but it's still using a heck of a lot of memory...
18252 jnasby 1 40 0 1135M 442M cpu1 290:03 48.07% postgres
19168 jnasby 1 60 0 62M 44M sleep 2:35 8.03% postgres
The second engine is typical of the other engines in terms of memory
usage. Also, I've seen some engines not releasing memory until
termination; there was one that was using 1.6G just sitting at the
prompt.
If it helps, that snapshot is with ~16M tuples in the table, according
to vacuum verbose (started with 0 tuples).
--
Jim C. Nasby (aka Decibel!) jim@nasby.net
Member: Triangle Fraternity, Sports Car Club of America
Give your computer some brain candy! www.distributed.net Team #1828
Windows: "Where do you want to go today?"
Linux: "Where do you want to go tomorrow?"
FreeBSD: "Are you guys coming, or what?"