Wes <wespvp(at)syntegra(dot)com> writes:
> It took 5.2 hours again tonight to do the vacuum.  I don't see anything out
> of the ordinary - no explanation for the non-linear increases in vacuum
> time.

Would you post the complete VACUUM VERBOSE log?  The CPU/elapsed time lines
would help us identify where the time is going.

> This is what shows up at the end:

> INFO:  free space map: 93 relations, 282 pages stored; 1712 total pages
> needed
> DETAIL:  Allocated FSM size: 1000 relations + 1000000 pages = 5920 kB shared
> memory.

Well, you don't have a problem with FSM being too small anyway ;-)

			regards, tom lane

In response to

• Re: Vacuum time degrading at 2005-03-02 15:46:46 from Wes

Responses

• Re: Vacuum time degrading at 2005-03-02 18:21:44 from Wes
• Re: Vacuum time degrading at 2005-03-02 18:36:01 from Wes
• Re: Vacuum time degrading at 2005-03-02 21:31:12 from Wes

pgsql-hackers by date

Next:	From: Wes	Date: 2005-03-02 18:21:44
	Subject: Re: Vacuum time degrading
Previous:	From: Tom Lane	Date: 2005-03-02 17:54:20
	Subject: Re: [HACKERS] UTF8 or Unicode

pgsql-general by date

Next:	From: Wes	Date: 2005-03-02 18:21:44
	Subject: Re: Vacuum time degrading
Previous:	From: Scott Frankel	Date: 2005-03-02 18:09:04
	Subject: Re: sql join question