On Sat, 2006-01-21 at 12:48 -0300, Alvaro Herrera wrote:
> Rod Taylor wrote:
> > Is there any way of bumping this limit or am I stuck breaking up the
> > transaction?
>
> Wow, I never heard of anyone reaching the limit :-( Sorry, you are
> stuck (short of changing CommandId to 64 bits, which would bloat your
> tables considerably ...)
Does a subtransaction receive an independent command counter? If so, I
could shove a bunch of work into large subtransactions allowing me to
avoid bumping the main transactions command counter.
Heck, PostgreSQL could automatically create a subtransaction (savepoint)
with no rollback name. The client sending a commit or rollback always
operates on the parent transaction and the child follows (committing or
rolling back when required).
--
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
