"Jim C. Nasby" <jnasby@pervasive.com> writes:
> What *might* make sense would be to provide two locations for pgsql_tmp,
> because a lot of operations in there involve reading and writing at the
> same time:
> Read from heap while writing tapes to pgsql_tmp
> read from tapes while writing final version to pgsql_tmp
Note that a large part of the reason for the current logtape.c design
is to avoid requiring 2X or more disk space to sort X amount of data.
AFAICS, any design that does the above will put us right back in the 2X
regime. That's a direct, measurable penalty; it'll take more than
handwaving arguments to convince me we should change it in pursuit of
unquantified speed benefits.
regards, tom lane