Re: Oracle vs PostgreSQL in real life - Mailing list pgsql-hackers

From Hannu Krosing
Subject Re: Oracle vs PostgreSQL in real life
Date
Msg-id 1014844117.2383.78.camel@rh72.home.ee
Whole thread Raw
In response to Re: Oracle vs PostgreSQL in real life  (Jean-Paul ARGUDO <jean-paul.argudo@idealx.com>)
List pgsql-hackers
On Wed, 2002-02-27 at 23:21, Jean-Paul ARGUDO wrote:
> > What was the postgresql.conf set to ?
> 
> I put parameters in another mail, please watch for it.
> 
> > > The "test" is a big batch that computes stuffs in the database. 
> > Could you run this batch in smaller chunks to see if PG is slow from the
> > start or does it slow down as it goes ?
> 
> The batch starts really fast and past 2 minuts, begins to slow down dramatically
> and never stops to get slower and slower
> 

I did a small test run on my home computer (Celeron 350, IDE disks,
untuned 7.2 on RH 7.2) 

I made a small table (int,text) with primary key on int and filled it
with values 1-512 for int.

then I ran a python script that updated 10000 random rows in patches of
10 updates.

the first run took 

a)   1.28 - 112 tps 

as it used seq scans

then I ran VACUUM ANALYZE and next runs were

1. 24 sec - 416 tps
2. 43 sec - 232 tps
3. 71 sec - 140 tps

then I tied the same query and run vacuum in another window manually
each 5 sec.

the result was similar to 1 - 24.5 sec

running vacuum every 10 sec slowed it to  25.1 sec, running every 3 sec
to 24.3 sec. Running vacuum in a tight loop slowed test down to 30.25
sec.


-------------------------
Hannu





pgsql-hackers by date:

Previous
From: Hannu Krosing
Date:
Subject: Re: Oracle vs PostgreSQL in real life
Next
From: Tom Lane
Date:
Subject: Re: Some additions/#ifdefs to target new OS NetWare