Horton

Re: Clock sweep not caching enough B-Tree leaf pages?

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Jim Nasby <Jim.Nasby@bluetreble.com>
Cc: Greg Stark <stark@mit.edu>, PostgreSQL-development <pgsql-hackers@postgresql.org>, Merlin Moncure <mmoncure@gmail.com>, Peter Geoghegan <pg@heroku.com>, Tom Lane <tgl@sss.pgh.pa.us>
Date: 2015-04-21T21:52:01Z
Lists: pgsql-hackers 
On Mon, Apr 20, 2015 at 2:53 PM, Jim Nasby <Jim.Nasby@bluetreble.com> wrote:
> I think that would help, but it still leaves user backends trying to advance
> the clock, which is quite painful. Has anyone tested running the clock in
> the background? We need a wiki page with all the ideas that have been tested
> around buffer management...

Amit's bgreclaimer patch did that, but we weren't able to demonstrate
a clear benefit.  I haven't given up on the idea yet, but we've got to
be able to prove that it's a good idea in practice as well as in
theory.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

Commits

  1. Replace the BufMgrLock with separate locks on the lookup hashtable and