pg.ddx.io  pgsql-admin@postgresql.org mailing list archive  
help / color / mirror / Atom feed
From: Scott Marlowe <scott.marlowe@gmail.com>
To: AmitKumar Jain <amitjain@synechron.com>
Cc: pgsql-admin@postgresql.org
Subject: Re: autovacuum benchmarking ...
Date: Thu, 4 Dec 2008 07:15:46 -0700
Message-ID: <dcc563d10812040615l75964715o8cbfb93cb7459fa2@mail.gmail.com> (raw)
In-Reply-To: <4937098C.4060605@synechron.com>
References: <4937098C.4060605@synechron.com>

On Wed, Dec 3, 2008 at 3:34 PM, AmitKumar Jain <amitjain@synechron.com> wrote:
> Hi Team
>
> 1. Can we know the how much time will take ongoing vacuum process  in DB?
> 2. Any benchmarking tool for vacuum process.
> 3. Can we know what exact operations has been done by Vacuum process.
> 4. I studied that only VACUUM process doesnot lock the table but due to it
> insertion in table gets very slow. when i killed vacuum daemon insertion get
> faster is it bug in 8.3.3
>
>
> I know I should not ask all these questions in one go but i need urgently
> its answer as vacuum process is just screwing my production database. I
> would thankful for any suggestion on above  question...

At least they're all related, and you're not spamming a half dozen
pgsql lists, so really, it's fine to put them all together.

1: Vacuum takes as long as it takes.  On faster machines it runs
faster.  Generally it's very I/O dependent, so lots of hard drives
will make it run faster.  It's also influenced by how much other
activity is going on.  Since it doesn't lock, it's not a big deal as
long as it finishes in a reasonable amount of time.
2: \timing
vacuum tablename;
3: logging, I believe you have to turn log_min_messages down (up?) to
get log entries. also look in pg_stat_user_tables.
4: You need to look into the settings autovacuum_vacuum_cost_delay and
vacuum_cost_delay.  Setting them to 10 or 20 or at most 30 or 40 will
slow down vacuum enough that it shouldn't too heavily impact the rest
of the system.



view thread (2+ messages)

Message-ID: <dcc563d10812040615l75964715o8cbfb93cb7459fa2@mail.gmail.com>
Permalink:  ../dcc563d10812040615l75964715o8cbfb93cb7459fa2@mail.gmail.com/
Also on:    postgresql.org/message-id/dcc563d10812040615l75964715o8cbfb93cb7459fa2@mail.gmail.com

 · 

reply

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Reply to all the recipients using the --to and --cc options:
  reply via email

  To: pgsql-admin@postgresql.org
  Cc: scott.marlowe@gmail.com, amitjain@synechron.com
  Subject: Re: autovacuum benchmarking ...
  In-Reply-To: <dcc563d10812040615l75964715o8cbfb93cb7459fa2@mail.gmail.com>

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox