Skip site navigation (1) Skip section navigation (2)

Re: speed up restore from dump

From: Sam Mason <sam(at)samason(dot)me(dot)uk>
To: pgsql-general(at)postgresql(dot)org
Subject: Re: speed up restore from dump
Date: 2008-10-30 20:01:56
Message-ID: 20081030200156.GB2459@frubble.xen.chris-lamb.co.uk (view raw or flat)
Thread:
Lists: pgsql-general
On Thu, Oct 30, 2008 at 07:28:57PM +0000, Joao Ferreira gmail wrote:
> On Thu, 2008-10-30 at 11:39 -0700, Alan Hodgson wrote: 
> > You're probably just 
> > disk-bound, though. What does vmstat say during the restore?
> 
> During restore:
> # vmstat
> procs --------memory------- ---swap-- -----io---- -system-- ----cpu----
> r  b   swpd   free   buff  cache   si  so   bi   bo  in  cs us sy id wa
> 3  1 230204   4972   1352 110128   2    1   17   63  24  56 12  2 85  0
> 
> 
> After restore has finished
> # vmstat
> procs --------memory-------- ---swap-- ---io---- -system-- ----cpu----
> r  b   swpd   free   buff  cache   si  so   bi   bo   in   cs us sy id wa
> 1  0 246864  59972   2276 186420   2   1    18   63   28   56 12  2 85  0

From the output you've given it doesn't look as though you left vmstat
running while the processing is running, the first set of numbers it
prints out are rarely representational values for the IO usage.  Try
running "vmstat 5" to get output every 5 seconds, you should be able
to see things happening a bit more easily that way.  Another tool I'd
recommend is iostat, I tend to invoke it as "iostat -mx 5 /dev/sd?" to
get it to print out values for each individual disk.


  Sam

In response to

pgsql-general by date

Next:From: Alan HodgsonDate: 2008-10-30 20:08:13
Subject: Re: speed up restore from dump
Previous:From: Joao Ferreira gmailDate: 2008-10-30 19:28:57
Subject: Re: speed up restore from dump

Privacy Policy | About PostgreSQL
Copyright © 1996-2014 The PostgreSQL Global Development Group