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

Re: can't handle large number of INSERT/UPDATEs

From: Curtis Zinzilieta <curtisz(at)norchemlab(dot)com>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: Anjan Dave <adave(at)vantage(dot)com>, Matt Clark <matt(at)ymogen(dot)net>,Rod Taylor <pg(at)rbt(dot)ca>,Postgresql Performance <pgsql-performance(at)postgresql(dot)org>
Subject: Re: can't handle large number of INSERT/UPDATEs
Date: 2004-10-26 23:04:35
Message-ID: Pine.LNX.4.44.0410261557560.23730-100000@hyperion2.norchemlab.com (view raw or flat)
Thread:
Lists: pgsql-performance
On Tue, 26 Oct 2004, Tom Lane wrote:

> "Anjan Dave" <adave(at)vantage(dot)com> writes:
> > One thing I am not sure is why 'bi' (disk writes) stays at 0 mostly,
> > it's the 'bo' column that shows high numbers (reads from disk). With so
> > many INSERT/UPDATEs, I would expect it the other way around...
> 
> Er ... it *is* the other way around.  bi is blocks in (to the CPU),
> bo is blocks out (from the CPU).
> 
> 			regards, tom lane

Ummm.....

[curtisz(at)labsoft T2]$ man vmstat
<snip>
FIELD DESCRIPTIONS
<snip>
   IO
       bi: Blocks sent to a block device (blocks/s).
       bo: Blocks received from a block device (blocks/s).

And on my read-heavy 7.4.2 system (running on rh8 at the moment....)
(truncated for readability...)

[root(at)labsoft T2]# vmstat 1
   procs                      memory    swap          io     system         
 r  b  w   swpd   free   buff  cache  si  so    bi    bo   in    cs  us  
 0  0  0 127592  56832 365496 2013788   0   1     3     6    4     0   4   
 2  0  0 127592  56868 365496 2013788   0   0     0     0  363   611   1   
 1  0  0 127592  57444 365508 2013788   0   0     8   972 1556  3616  11  
 0  0  1 127592  57408 365512 2013800   0   0     0   448  614  1216   5   
 0  0  0 127592  56660 365512 2013800   0   0     0     0  666  1150   6   
 0  3  1 127592  56680 365512 2013816   0   0    16   180 1280  2050   2   
 0  0  0 127592  56864 365516 2013852   0   0    20   728 2111  4360  11   
 0  0  0 127592  57952 365544 2013824   0   0     0   552 1153  2002  10   
 0  0  0 127592  57276 365544 2013824   0   0     0   504  718  1111   5   
 1  0  0 127592  57244 365544 2013824   0   0     0   436 1495  2366   7   
 0  0  0 127592  57252 365544 2013824   0   0     0     0  618  1380   5   
 0  0  0 127592  57276 365556 2014192   0   0   360  1240 2418  5056  14   
 2  0  0 127592  56664 365564 2014176   0   0     0   156  658  1349   5   
 1  0  0 127592  55864 365568 2014184   0   0     0  1572 1388  3598   9   
 2  0  0 127592  56160 365572 2014184   0   0     0   536 4860  6621  13   

Which seems appropriate for both the database and the man page....

-Curtis



In response to

Responses

pgsql-performance by date

Next:From: Thomas F.O'ConnellDate: 2004-10-27 00:09:22
Subject: Re: [PATCHES] [HACKERS] ARC Memory Usage analysis
Previous:From: Anjan DaveDate: 2004-10-26 22:45:37
Subject: Re: can't handle large number of INSERT/UPDATEs

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