Обсуждение: suppress output for benchmarking

 I am testing the performance of postgresql on a set of workloads. However,<br /> the output significantly affects the
performanceevaluation. Is there a way<br /> to by-pass all output of select statements so the timing reflects only the
<br/> query evaluation process?<br /><br /> thanks a lot<br /> 

On Wed, Jan 18, 2006 at 10:35:48PM -0500, uwcssa wrote:
>  I am testing the performance of postgresql on a set of workloads. However,
> the output significantly affects the performance evaluation. Is there a way
> to by-pass all output of select statements so the timing reflects only the
> query evaluation process?

SELECT count(*) FROM (SELECT ...) a;

If you're using psql \timing will probably be useful as well.

And this is better suited for -general...
-- 
Jim C. Nasby, Sr. Engineering Consultant      jnasby@pervasive.com
Pervasive Software      http://pervasive.com    work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf       cell: 512-569-9461

Ühel kenal päeval, K, 2006-01-18 kell 22:35, kirjutas uwcssa:
>  I am testing the performance of postgresql on a set of workloads.
> However,
> the output significantly affects the performance evaluation. Is there
> a way
> to by-pass all output of select statements so the timing reflects only
> the 
> query evaluation process?

If you do EXPLAIN ANALYSE QUERY instead of just QUERY, then the backend
discards all rows returned and just gives back performance data

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

On Thu, Jan 19, 2006 at 11:46:29PM +0200, Hannu Krosing wrote:
> ??hel kenal p??eval, K, 2006-01-18 kell 22:35, kirjutas uwcssa:
> >  I am testing the performance of postgresql on a set of workloads.
> > However,
> > the output significantly affects the performance evaluation. Is there
> > a way
> > to by-pass all output of select statements so the timing reflects only
> > the 
> > query evaluation process?
> 
> If you do EXPLAIN ANALYSE QUERY instead of just QUERY, then the backend
> discards all rows returned and just gives back performance data

The flipside is that EXPLAIN ANALYZE adds it's own (non-trivial)
overhead to the query.
-- 
Jim C. Nasby, Sr. Engineering Consultant      jnasby@pervasive.com
Pervasive Software      http://pervasive.com    work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf       cell: 512-569-9461

Fellow hacker,

I am new comer to postgres development community. Currently, I am 
implementing tightly coupled machine classifiers within postgres. The 
grammer looks like Train Parameter_list (class1,class2,class3...). I have 
two major problems right now.

1. Train is a statement and it is suppose to return some parameters in the 
form of a query. (To be used by a classifier later.) How can I return a 
Query with self specified column name and data vectors?

2. class1, class2 are all supposed to be relations/query result. But what is 
a proper container to hold them. I found var_list to be a good candidate as 
it can contain unconstrained number of var_values. But var_values are of 
constant types. Can I just add query as a constant? Anyone had this kind of 
implementing experience?

Thanks,
John