On Monday, January 28, 2013, Kevin Grittner wrote:
IMO, anything which changes an anti-wraparound vacuum of a
bulk-loaded table from "read the entire table and rewrite nearly
the complete table with WAL-logging" to rewriting a smaller portion
of the table with WAL-logging is an improvement. Anyone who has
run an OLTP load on a database which was loaded from pg_dump output
or other bulk load processes, has probably experienced the pain
related to the WAL-logged rewrite of massive quantities of data.
pgbench seems to be the OLTP load par excellence (or perhaps ad nauseum).
What other set up is needed to get it to reproduce this problem? Do we just do a dump/restore in lieu of pgbench -i?
Cheers,
Jeff