Re: Problems with vacuum!

From: "Chris Hoover" <revoohc(at)sermonaudio(dot)com>
To: "Domenico Sgarbossa" <domenico(at)xtecnica(dot)com>
Cc: <pgsql-performance(at)postgresql(dot)org>
Subject: Re: Problems with vacuum!
Date: 2004-06-11 14:39:59
Message-ID: 40C9C43F.2060402@sermonaudio.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-performance

Domenico Sgarbossa wrote:

> I'm running postgrSQL 7.2 on a linux Red Hat 8.0 box with 2GB of RAM
> When I boot-up the system , this is the TOP situation:
>
> 11:59am up 4 min, 1 user, load average: 0.37, 0.26, 0.11
>
> 77 processes: 74 sleeping, 3 running, 0 zombie, 0 stopped
>
> CPU states: 0.3% user, 0.7% system, 0.0% nice, 98.8% idle
>
> Mem: 1031020K av, 177808K used, 853212K free, 0K shrd,
> 14744K buff
>
> Swap: 2096472K av, 0K used, 2096472K free
> 67828K cached
>
> After I've done a vacuum , the situation is:
>
>
>
> 12:04pm up 8 min, 1 user, load average: 0.22, 0.23, 0.12
>
> 78 processes: 76 sleeping, 2 running, 0 zombie, 0 stopped
>
> CPU states: 2.5% user, 1.9% system, 0.0% nice, 95.4% idle
>
> Mem: 1031020K av, 1016580K used, 14440K free, 0K shrd,
> 18624K buff
>
> Swap: 2096472K av, 0K used, 2096472K free
> 833896K cached
>
> As you see the memory used by vacuum isn't released anymore? Anyone
> know why?
>
> My statement is: vacuumdb --analyze dbname
> My pg paramaters on postgresql.conf are:
>
> #
> # Shared Memory Size
> #
> shared_buffers = 2048 # 2*max_connections, min 16
> max_fsm_relations = 100 # min 10, fsm is free space map
> max_fsm_pages = 10000 # min 1000, fsm is free space map
> max_locks_per_transaction = 64 # min 10
> wal_buffers = 8 # min 4
>
> #
> # Non-shared Memory Sizes
> #
> sort_mem = 512 # min 32
> vacuum_mem = 8192 # min 1024
> Did anyone know why this happend?
>
> Distinti Saluti
>
> Sgarbossa Domenico
> X Tecnica S.R.L.
> www.xtecnica.com <http://www.xtecnica.com>
> Tel: 049/9409154 - 049/5970297
> Fax: 049/9400288

Your vacuum memory was released. However, the linux kernel likes to
keep most memory in the cached buffer instead of totally freeing it.
The guru's behind the kernel figured out they get better performance
this way.

HTH,

Chris

In response to

Browse pgsql-performance by date

  From Date Subject
Next Message Domenico Sgarbossa 2004-06-11 14:53:22 Re: [BULK] Problems with vacuum!
Previous Message Domenico Sgarbossa 2004-06-11 13:54:44 Problems with vacuum!