"Marc G. Fournier" <scrappy(at)postgresql(dot)org> writes:
> On Thu, 11 Aug 2005, Tom Lane wrote:
>> If you've got any AFTER UPDATE triggers on that table, you could be
>> running out of memory for the pending-triggers list.
> Nope, only have a BEFORE UPDATE, or would that be similar except for at
> which point it runs out of memory?
Nope, BEFORE UPDATE shouldn't result in any permanent memory
accumulation.
An out-of-memory error should result in a long report in the postmaster
log about how many bytes in each memory context --- can you post that?
regards, tom lane