While examining pg_upgrade on a cluster with many tables (created with the
command in [0]), I noticed that a huge amount of pg_dump time goes towards
the binary_upgrade_set_pg_class_oids() function. This function executes a
rather expensive query for a single row, and this function appears to be
called for most of the rows in pg_class.
The attached work-in-progress patch speeds up 'pg_dump --binary-upgrade'
for this case. Instead of executing the query in every call to the
function, we can execute it once during the first call and store all the
required information in a sorted array that we can bsearch() in future
calls. For the aformentioned test, pg_dump on my machine goes from ~2
minutes to ~18 seconds, which is much closer to the ~14 seconds it takes
without --binary-upgrade.
One downside of this approach is the memory usage. This was more-or-less
the first approach that crossed my mind, so I wouldn't be surprised if
there's a better way. I tried to keep the pg_dump output the same, but if
that isn't important, maybe we could dump all the pg_class OIDs at once
instead of calling binary_upgrade_set_pg_class_oids() for each one.
[0] https://postgr.es/m/3612876.1689443232%40sss.pgh.pa.us
--
Nathan Bossart
Amazon Web Services: https://aws.amazon.com