Re: Index bloat of 4x

Поиск
Список
Период
Сортировка
Искать
От
Bruce Momjian
Тема
Re: Index bloat of 4x
Дата
Msg-id
200701310415.l0V4FtP19125@momjian.us
Ответ на
Список
Дерево обсуждения
Index bloat of 4x Bill Moran <wmoran@collaborativefusion.com>
Re: Index bloat of 4x "Greg Sabino Mullane" <greg@turnstep.com>
Re: Index bloat of 4x Csaba Nagy <nagy@ecircle-ag.com>
Re: Index bloat of 4x Alvaro Herrera <alvherre@commandprompt.com>
Re: Index bloat of 4x Bill Moran <wmoran@collaborativefusion.com>
Re: Index bloat of 4x Alvaro Herrera <alvherre@commandprompt.com>
Re: Index bloat of 4x Bill Moran <wmoran@collaborativefusion.com>
Re: Index bloat of 4x Tom Lane <tgl@sss.pgh.pa.us>
Re: Index bloat of 4x Bill Moran <wmoran@collaborativefusion.com>
Re: Index bloat of 4x "Ed L." <pgsql@bluepolka.net>
Re: Index bloat of 4x Csaba Nagy <nagy@ecircle-ag.com>
Re: Index bloat of 4x "Ed L." <pgsql@bluepolka.net>
Re: Index bloat of 4x Tom Lane <tgl@sss.pgh.pa.us>
Re: Index bloat of 4x "Jeremy Haile" <jhaile@fastmail.fm>
Re: Index bloat of 4x Csaba Nagy <nagy@ecircle-ag.com>
Re: Index bloat of 4x "Ed L." <pgsql@bluepolka.net>
Re: Index bloat of 4x Bruce Momjian <bruce@momjian.us>
Re: Index bloat of 4x Jim Nasby <decibel@decibel.org>
Re: Index bloat of 4x Alvaro Herrera <alvherre@commandprompt.com>
Re: Index bloat of 4x Bruce Momjian <bruce@momjian.us>
Re: Index bloat of 4x Tom Lane <tgl@sss.pgh.pa.us>
Re: Index bloat of 4x Bruce Momjian <bruce@momjian.us>
Re: Index bloat of 4x Stefan Kaltenbrunner <stefan@kaltenbrunner.cc>
Re: Index bloat of 4x Vivek Khera <vivek@khera.org>
Tom Lane wrote:
> Bill Moran  writes:
> > The entire database was around 28M prior to the upgrades, etc.  Immediately
> > after the upgrades, it was ~270M.  Following a vacuum full, it dropped to
> > 165M.  Following a database-wide reindex, it dropped to 30M.
> 
> As Alvaro said, vacuum full doesn't shrink indexes but in fact bloats them.
> (Worst case, they could double in size, if the vacuum moves every row;
> there's an intermediate state where there have to be index entries for
> both old and new copies of each moved row, to ensure things are
> consistent if the vacuum crashes right there.)
> 
> So the above doesn't sound too unlikely.  Perhaps we should recommend
> vac full + reindex as standard cleanup procedure.  Longer term, maybe
> teach vac full to do an automatic reindex if it's moved more than X% of
> the rows.  Or forget the current vac full implementation entirely, and
> go over to something acting more like CLUSTER ...

TODO already has:

	* Improve speed with indexes
	
	  For large table adjustments during VACUUM FULL, it is faster to
	  reindex rather than update the index.  Also, index updates can
	  bloat the index.

-- 
  Bruce Momjian   bruce@momjian.us
  EnterpriseDB    http://www.enterprisedb.com

  + If your life is a hard drive, Christ can be your backup. +
В списке pgsql-general по дате отправления
От: Tom Lane
Дата:
От: David Fetter
Дата:
FAQ