Re: Removing freelist (was Re: Should I implement DROP INDEX CONCURRENTLY?)
Robert Haas <robertmhaas@gmail.com>
From: Robert Haas <robertmhaas@gmail.com>
To: Jim Nasby <jim@nasby.net>
Cc: Heikki Linnakangas <heikki.linnakangas@enterprisedb.com>, Simon Riggs <simon@2ndquadrant.com>, Tom Lane <tgl@sss.pgh.pa.us>, Daniel Farina <daniel@heroku.com>, Merlin Moncure <mmoncure@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2012-01-23T02:06:39Z
Lists: pgsql-hackers
On Sat, Jan 21, 2012 at 5:29 PM, Jim Nasby <jim@nasby.net> wrote: > We should also look at having the freelist do something useful, instead of just dropping it completely. Unfortunately that's probably more work... That's kinda my feeling as well. The free list in its current form is pretty much useless, but I don't think we'll save much by getting rid of it, because that's just a single test. The expensive part of what we do while holding BufFreelistLock is, I think, iterating through buffers taking and releasing a spinlock on each one (!). So I guess my vote would be to leave it alone pending further study, and maybe remove it later if we can't find a way to rejigger it to be more useful. -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company