pg.ddx.io  pgsql-php@postgresql.org mailing list archive  
help / color / mirror / Atom feed
Long running queries and timeouts
16+ messages / 11 participants
[nested] [flat]

* Long running queries and timeouts
@ 2002-01-19 16:48 ` Francisco Reyes <lists@natserv.com>
  2002-01-19 17:32   ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
  1 sibling, 1 reply; 16+ messages in thread

From: Francisco Reyes @ 2002-01-19 16:48 UTC (permalink / raw)
  To: pgsql-php

New to the PHP/PostgreSQL combo.
I have a couple of queries which take a few minutes to run and we are
getting timeouts. Didn't find anything usefull on the archives.

How are you handling long running queries on your PHP?




^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-19 16:48 ` Long running queries and timeouts Francisco Reyes <lists@natserv.com>
@ 2002-01-19 17:32   ` Chadwick Rolfs <cmr@shell.gis.net>
  2002-01-20 00:10     ` Re: Long running queries and timeouts Mitch Vincent <mitch@doot.org>
  2002-01-20 15:56     ` Re: Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  0 siblings, 2 replies; 16+ messages in thread

From: Chadwick Rolfs @ 2002-01-19 17:32 UTC (permalink / raw)
  To: Francisco Reyes <lists@natserv.com>; +Cc: pgsql-php

On Sat, 19 Jan 2002, Francisco Reyes wrote:

> New to the PHP/PostgreSQL combo.
> I have a couple of queries which take a few minutes to run and we are
> getting timeouts. Didn't find anything usefull on the archives.
> 
> How are you handling long running queries on your PHP?
> 
> 
> ---------------------------(end of broadcast)---------------------------
> TIP 2: you can get off all lists at once with the unregister command
>     (send "unregister YourEmailAddressHere" to majordomo@postgresql.org)
> 
You can always use persistent connections.  It's not a good idea to use
too many, or you will be denied access to Postgres any other way, but it's
a quick fix for the short run.

Long run and responsibility includes using an array, or an object to
retrieve data.  You can always split queries up into many small queries.
We'll need a little more information to help you with that, though.  Like
what the code is...what your php.ini file and httpd.conf file says about
connection timeouts...  maybe more ??-- like what you've read on php.net's
docs...

When new programmers here where I work started using php/postgresql 6
months ago, they used pconnect in EVERY one of their scripts, filled up
the server, and were unable to connect.  No one was able to.  They didn't
know what they did, and it took a little reading to figure out that Apache
needs to be restarted, NOT postgres.  Don't kill -9 the postmaster.
Regards,

Chadwick Rolfs - cmr@gis.net
Cleveland State University - Student
Music Major - The Holden Arboretum Volunteer
Computer Programmer - Student Employee
--*I finally found powdered water;
I just can't figure out what to add to it*--




^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-19 16:48 ` Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  2002-01-19 17:32   ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
@ 2002-01-20 00:10     ` Mitch Vincent <mitch@doot.org>
  2002-01-20 02:59       ` Re: Long running queries and timeouts Cornelia Boenigk <poppcorn@cornelia-boenigk.de>
  2002-01-21 17:02       ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
  1 sibling, 2 replies; 16+ messages in thread

From: Mitch Vincent @ 2002-01-20 00:10 UTC (permalink / raw)
  To: pgsql-php

http://www.php.net/manual/en/function.set-time-limit.php will allow you to
set the timeout but you might post the queries and EXPLAIN output to
pg-general to see if there is something that can be done to speed them up...
Some well placed indexes always help!

> You can always use persistent connections.  It's not a good idea to use
> too many, or you will be denied access to Postgres any other way, but it's
> a quick fix for the short run.

The maximum connections is a configuration parameter in postgresql.conf --
persistent connections are pretty cool but have been broken until recent
versions of PHP so I'm still hesitant using them..

> Long run and responsibility includes using an array, or an object to
> retrieve data.

Huh?

>You can always split queries up into many small queries.

What makes you say that?

-Mitch





^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-19 16:48 ` Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  2002-01-19 17:32   ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
  2002-01-20 00:10     ` Re: Long running queries and timeouts Mitch Vincent <mitch@doot.org>
@ 2002-01-20 02:59       ` Cornelia Boenigk <poppcorn@cornelia-boenigk.de>
  2002-01-20 14:56         ` Re: Long running queries and timeouts jeff fitzmyers <fitz22@earthlink.net>
  1 sibling, 1 reply; 16+ messages in thread

From: Cornelia Boenigk @ 2002-01-20 02:59 UTC (permalink / raw)
  To: Mitch Vincent <mitch@doot.org>; pgsql-php

Hi all

> http://www.php.net/manual/en/function.set-time-limit.php will
allow you to
> set the timeout

This only works if safe mode is off.

If you have a couple of queries in one script you can divide them
up on different pages. Put a redirect at the end of the query. On
the new page the execution time starts new. Normally it is set to
30 seconds in the php.ini file.
Or have a look on the function ini_set()

string ini_set (string varname, string newvalue)

(I have never tested this function;-)

Greetings
Conni







^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-19 16:48 ` Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  2002-01-19 17:32   ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
  2002-01-20 00:10     ` Re: Long running queries and timeouts Mitch Vincent <mitch@doot.org>
  2002-01-20 02:59       ` Re: Long running queries and timeouts Cornelia Boenigk <poppcorn@cornelia-boenigk.de>
@ 2002-01-20 14:56         ` jeff fitzmyers <fitz22@earthlink.net>
  0 siblings, 0 replies; 16+ messages in thread

From: jeff fitzmyers @ 2002-01-20 14:56 UTC (permalink / raw)
  To: pgsql-php

> If you have a couple of queries in one script you can divide them
> up on different pages.

I have a page that uses many queries to gather a bunch of stats that are 
saved for later display. It is not too important that they be noted at 
exactly the same time so every xxth time (or after a timestamp)  1 of 
them is updated. The next case to update is saved in the row. Adds a bit 
of overhead, but I don't have to worry as the db grows and can easily 
add more cases.

Jeff




^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-19 16:48 ` Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  2002-01-19 17:32   ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
  2002-01-20 00:10     ` Re: Long running queries and timeouts Mitch Vincent <mitch@doot.org>
@ 2002-01-21 17:02       ` Chadwick Rolfs <cmr@shell.gis.net>
  2002-01-22 01:30         ` Re: Long running queries and timeouts Christopher Kings-Lynne <chriskl@familyhealth.com.au>
  1 sibling, 1 reply; 16+ messages in thread

From: Chadwick Rolfs @ 2002-01-21 17:02 UTC (permalink / raw)
  To: pgsql-php

On Sat, 19 Jan 2002, Mitch Vincent wrote:

> http://www.php.net/manual/en/function.set-time-limit.php will allow you to
> set the timeout but you might post the queries and EXPLAIN output to
> pg-general to see if there is something that can be done to speed them up...
> Some well placed indexes always help!
> 
> > You can always use persistent connections.  It's not a good idea to use
> > too many, or you will be denied access to Postgres any other way, but it's
> > a quick fix for the short run.
> 
> The maximum connections is a configuration parameter in postgresql.conf --
> persistent connections are pretty cool but have been broken until recent
> versions of PHP so I'm still hesitant using them..
> 

It's good to see someone else sees problems with persistent connections.
You could set your maximum connections to infinite, but would that really
do any good?  It seems the server might easily run out of memory if there
isn't a cap on connection numbers,,,, as well as connection time.

> > Long run and responsibility includes using an array, or an object to
> > retrieve data.
> 
> Huh?

I can see why this doesn't make much sense.  Seeing as we have nothing to
work with from the original post regarding just what the query is.. I'll
try to give an example to explain.

querying all the names in a phone book DB:
select first_name, last_name from white_pages;
--UH... this will be,,,let's say 150,000 entries (small city phone book)

WELL, you could select by starting letters, and put it into an array,
using pg_fetch_array, or pg_fetch_object.  Make multiple queries in the
same script, and use that array (or object) to ouput to the browser...

select first_name, last_name from white_pages where last_name like 'A%';
select first_name, last_name from white_pages where last_name like 'B%';
select first_name, last_name from white_pages where last_name like 'C%';
select first_name, last_name from white_pages where last_name like 'D%';
...etc

if it's still too long, try 'Aa%', 'Ab%' ... so on.  Tedious, but
efficient.

I know this is only the sql, there is php involved.  We can save that for
the php list, or, if there are any further questions.  Hope this clarifies
at least a little bit. ??

> 
> >You can always split queries up into many small queries.
> 
> What makes you say that?

if you have your rdbms set up right, you have a unique number (primary
key) relating to each row.  So using the sql syntax above, you could
select one row at a time, if you needed to, using a for loop and sending
single queries to the backend.  You shouldn't be flooding a webbrowser
with over 30 seconds of data anyway, as you run the risk of a crash.

> 
> -Mitch
> 
> 
> 
> ---------------------------(end of broadcast)---------------------------
> TIP 6: Have you searched our list archives?
> 
> http://archives.postgresql.org
> 

Chadwick Rolfs - cmr@gis.net
Cleveland State University - Student
Music Major - The Holden Arboretum Volunteer
Computer Programmer - Student Employee
--*I finally found powdered water;
I just can't figure out what to add to it*--




^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-19 16:48 ` Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  2002-01-19 17:32   ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
  2002-01-20 00:10     ` Re: Long running queries and timeouts Mitch Vincent <mitch@doot.org>
  2002-01-21 17:02       ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
@ 2002-01-22 01:30         ` Christopher Kings-Lynne <chriskl@familyhealth.com.au>
  0 siblings, 0 replies; 16+ messages in thread

From: Christopher Kings-Lynne @ 2002-01-22 01:30 UTC (permalink / raw)
  To: Chadwick Rolfs <cmr@shell.gis.net>; pgsql-php

> I can see why this doesn't make much sense.  Seeing as we have nothing to
> work with from the original post regarding just what the query is.. I'll
> try to give an example to explain.
>
> querying all the names in a phone book DB:
> select first_name, last_name from white_pages;
> --UH... this will be,,,let's say 150,000 entries (small city phone book)
>
> WELL, you could select by starting letters, and put it into an array,
> using pg_fetch_array, or pg_fetch_object.  Make multiple queries in the
> same script, and use that array (or object) to ouput to the browser...
>
> select first_name, last_name from white_pages where last_name like 'A%';
> select first_name, last_name from white_pages where last_name like 'B%';
> select first_name, last_name from white_pages where last_name like 'C%';
> select first_name, last_name from white_pages where last_name like 'D%';
> ...etc
>
> if it's still too long, try 'Aa%', 'Ab%' ... so on.  Tedious, but
> efficient.

You could always just use CURSORs.  Look up 'DECLARE' and 'MOVE' and 'FETCH'
sql commands in postgres.

Chris




^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-19 16:48 ` Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  2002-01-19 17:32   ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
@ 2002-01-20 15:56     ` Francisco Reyes <lists@natserv.com>
  2002-01-21 10:21       ` Re: Long running queries and timeouts Marco Colombo <marco@esi.it>
  1 sibling, 1 reply; 16+ messages in thread

From: Francisco Reyes @ 2002-01-20 15:56 UTC (permalink / raw)
  To: Chadwick Rolfs <cmr@shell.gis.net>; +Cc: pgsql-php

On Sat, 19 Jan 2002, Chadwick Rolfs wrote:

> On Sat, 19 Jan 2002, Francisco Reyes wrote:
>
> > New to the PHP/PostgreSQL combo.
> > I have a couple of queries which take a few minutes to run and we are
> > getting timeouts. Didn't find anything usefull on the archives.
> >
> > How are you handling long running queries on your PHP?
> >
> You can always use persistent connections.  It's not a good idea to use
> too many, or you will be denied access to Postgres any other way, but it's
> a quick fix for the short run.
>
> Long run and responsibility includes using an array, or an object to
> retrieve data.  You can always split queries up into many small queries.
> We'll need a little more information to help you with that, though.  Like
> what the code is...what your php.ini file and httpd.conf file says about
> connection timeouts...  maybe more ??-- like what you've read on php.net's
> docs...

Thanks for the info/pointers. The issue is basically a query that takes
several minutes to run. I look/read about persistent connections and I
don't think that has anything to do with the problem. I believe the
timeout is probably at the web server.

After someone suggested I looked at "set_time_limit" I found that such
function doesn't account for DBconnection time. Afer reading that and your
suggestion to read httpd.conf I am starting to think that the likely place
of the timeout is the web server.

Will play with the settings on httpd.conf

I wonder how sites that display an update while the long query is running
do it. I have seen sites that display something like "retrieving info..
please wait" and then right before the info comes up the waiting text
dissapears. Is that Javascript? Flash? I can't imagine how that could be
done with plain html.




^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-19 16:48 ` Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  2002-01-19 17:32   ` Re: Long running queries and timeouts Chadwick Rolfs <cmr@shell.gis.net>
  2002-01-20 15:56     ` Re: Long running queries and timeouts Francisco Reyes <lists@natserv.com>
@ 2002-01-21 10:21       ` Marco Colombo <marco@esi.it>
  0 siblings, 0 replies; 16+ messages in thread

From: Marco Colombo @ 2002-01-21 10:21 UTC (permalink / raw)
  To: Francisco Reyes <lists@natserv.com>; +Cc: pgsql-php

On Sunday 20 January 2002 16:56, Francisco Reyes wrote:
> On Sat, 19 Jan 2002, Chadwick Rolfs wrote:
> > On Sat, 19 Jan 2002, Francisco Reyes wrote:
> > > New to the PHP/PostgreSQL combo.
> > > I have a couple of queries which take a few minutes to run and we are
> > > getting timeouts. Didn't find anything usefull on the archives.
> > >
> > > How are you handling long running queries on your PHP?
[...]
> After someone suggested I looked at "set_time_limit" I found that such
> function doesn't account for DBconnection time. Afer reading that and your
> suggestion to read httpd.conf I am starting to think that the likely place
> of the timeout is the web server.
>
> Will play with the settings on httpd.conf
>
> I wonder how sites that display an update while the long query is running
> do it. I have seen sites that display something like "retrieving info..
> please wait" and then right before the info comes up the waiting text
> dissapears. Is that Javascript? Flash? I can't imagine how that could be
> done with plain html.

What timeout message do you get, exactly? I could be just the *client*, or
the HTTP proxy, that times out if your script doesn't produce any output,
and *not* the httpd process (thus, no need to look at httpd.conf).

.TM.



^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
@ 2002-01-19 17:42 ` Frank Bax <fbax@sympatico.ca>
  2002-01-20 15:50   ` Re: Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  1 sibling, 1 reply; 16+ messages in thread

From: Frank Bax @ 2002-01-19 17:42 UTC (permalink / raw)
  To: Francisco Reyes <lists@natserv.com>; +Cc: pgsql-php


	set_time_limit( <secs> );


At 11:48 AM 1/19/02 -0500, Francisco Reyes wrote:
>New to the PHP/PostgreSQL combo.
>I have a couple of queries which take a few minutes to run and we are
>getting timeouts. Didn't find anything usefull on the archives.
>
>How are you handling long running queries on your PHP?
>
>
>---------------------------(end of broadcast)---------------------------
>TIP 2: you can get off all lists at once with the unregister command
>    (send "unregister YourEmailAddressHere" to majordomo@postgresql.org)
>



^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-19 17:42 ` Re: Long running queries and timeouts Frank Bax <fbax@sympatico.ca>
@ 2002-01-20 15:50   ` Francisco Reyes <lists@natserv.com>
  0 siblings, 0 replies; 16+ messages in thread

From: Francisco Reyes @ 2002-01-20 15:50 UTC (permalink / raw)
  To: Frank Bax <fbax@sympatico.ca>; +Cc: pgsql-php

On Sat, 19 Jan 2002, Frank Bax wrote:

>
> 	set_time_limit( <secs> );
>
>
> At 11:48 AM 1/19/02 -0500, Francisco Reyes wrote:
> >How are you handling long running queries on your PHP?

I just read the explanation at the php.net site for that function and it
doesn't seem to do what I would need:
"Any time spent on activity that happens outside the execution of the
script  such as ......database queries, etc. is not included when
determining the maximum time that the script has been running."

I am going to try anyway to see if this helps.




^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
@ 2002-01-22 15:52 Josh Berkus <josh@agliodbs.com>
  2002-01-23 03:54 ` Re: Long running queries and timeouts Francisco Reyes <lists@natserv.com>
  2002-01-23 13:58 ` Re: Long running queries and timeouts Jean-Michel POURE <jm.poure@freesurf.fr>
  0 siblings, 2 replies; 16+ messages in thread

From: Josh Berkus @ 2002-01-22 15:52 UTC (permalink / raw)
  To: Francisco Reyes <lists@natserv.com>; +Cc: pgsql-php

Francisco,

> > If queries > 2 min are common, you'll need some kind of "progress
>  bar"
> >  mechanism that updates the page every 40 seconds to prevent the
> >  browser from timing out.
> > -Josh
> 
> Do you have any experience on this? Sample code?
> Is that a term you think I may search for on the web "progress bar"?

Actually, my web developer and I talked about the programming necessary
 to do this, and decided that it was easier to break up the
 long-running function into 3 smaller functions.  So, no, we don't have
 any sample code.  In fact, we might be willing to buy some code that
 creates a progress bar from someone who does have it ...

-Josh Berkus

______AGLIO DATABASE SOLUTIONS___________________________
                                       Josh Berkus
  Complete information technology      josh@agliodbs.com
   and data management solutions       (415) 565-7293
  for law firms, small businesses        fax 621-2533
    and non-profit organizations.      San Francisco



^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-22 15:52 Re: Long running queries and timeouts Josh Berkus <josh@agliodbs.com>
@ 2002-01-23 03:54 ` Francisco Reyes <lists@natserv.com>
  2002-01-23 20:39   ` Re: Long running queries and timeouts postgresql@finner.de
  1 sibling, 1 reply; 16+ messages in thread

From: Francisco Reyes @ 2002-01-23 03:54 UTC (permalink / raw)
  To: Josh Berkus <josh@agliodbs.com>; +Cc: pgsql-php

On Tue, 22 Jan 2002, Josh Berkus wrote:

> Actually, my web developer and I talked about the programming necessary
>  to do this, and decided that it was easier to break up the
>  long-running function into 3 smaller functions.  So, no, we don't have
>  any sample code.  In fact, we might be willing to buy some code that
>  creates a progress bar from someone who does have it ...

I can't imagine a "progress bar". In particular if you are waiting for a
reply from the DB server.




^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-22 15:52 Re: Long running queries and timeouts Josh Berkus <josh@agliodbs.com>
  2002-01-23 03:54 ` Re: Long running queries and timeouts Francisco Reyes <lists@natserv.com>
@ 2002-01-23 20:39   ` postgresql@finner.de
  0 siblings, 0 replies; 16+ messages in thread

From: postgresql@finner.de @ 2002-01-23 20:39 UTC (permalink / raw)
  To: Francisco Reyes <lists@natserv.com>; +Cc: pgsql-php; Josh Berkus <josh@agliodbs.com>


On 23-Jan-02 Francisco Reyes sat down, thought for a long time and then
wrote:
> On Tue, 22 Jan 2002, Josh Berkus wrote:
> 
>> Actually, my web developer and I talked about the programming
>> necessary
>>  to do this, and decided that it was easier to break up the
>>  long-running function into 3 smaller functions.  So, no, we don't
>>  have
>>  any sample code.  In fact, we might be willing to buy some code
>>  that
>>  creates a progress bar from someone who does have it ...
> 
> I can't imagine a "progress bar". In particular if you are waiting
> for a
> reply from the DB server.

I found a nice one on http://www.lyrics.de! ;-)

It´s just a fake: An animated GIF picture with a progress bar. I think,
they made some statistics on database queries, added 10% and created
the GIF. The result is, that the waiting time seems shorter. Nice.

Correct me anybody please, who can proof that this progress bar is no
fake. I´d like to know this technique, too.

Greetings,
--
Frank Finner

And now there is no turning back at all.
                              (M. Moorcock, "Elric Of Melnibone")"



^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-22 15:52 Re: Long running queries and timeouts Josh Berkus <josh@agliodbs.com>
@ 2002-01-23 13:58 ` Jean-Michel POURE <jm.poure@freesurf.fr>
  2002-01-26 00:27   ` Re: Long running queries and timeouts Josh Berkus <josh@agliodbs.com>
  1 sibling, 1 reply; 16+ messages in thread

From: Jean-Michel POURE @ 2002-01-23 13:58 UTC (permalink / raw)
  To: Josh Berkus <josh@agliodbs.com>; Francisco Reyes <lists@natserv.com>; +Cc: pgsql-php

Le Mardi 22 Janvier 2002 16:52, Josh Berkus a écrit :
> Actually, my web developer and I talked about the programming necessary
>  to do this, and decided that it was easier to break up the
>  long-running function into 3 smaller functions.

Are you connecting from Netscape? I noticed long queries timeouted in 
Netscape. This is because Nescape waits for the entire content of a page 
before displaying it.

On the converse, when connecting from $M Internet Explorer, just echo a 
message like "Please wait during processing." and then "." everytime you run 
a query. IE refreshes content and there is no timeout.

Also, it is possible to set timeout time in /etc/php.ini. Set a high value 
like 600 seconds. Then you also need to allow more connections to PostgreSQL 
backend as described in this very interesting article : 
http://www.phpbuilder.com/columns/smith20010821.php3?page=2.

PostgreSQL can run very fast when well-optimized. You probably need to 
optimize the source code of queries themselves, probably writing some good 
PL/pgSQL server side code. If you need a GUI to do so, download and install 
Dave Page's excellent tool http://pgadmin.postgresql.org

Cheers,
Jean-Michel POURE



^ permalink  raw  reply  [nested|flat] 16+ messages in thread

* Re: Long running queries and timeouts
  2002-01-22 15:52 Re: Long running queries and timeouts Josh Berkus <josh@agliodbs.com>
  2002-01-23 13:58 ` Re: Long running queries and timeouts Jean-Michel POURE <jm.poure@freesurf.fr>
@ 2002-01-26 00:27   ` Josh Berkus <josh@agliodbs.com>
  0 siblings, 0 replies; 16+ messages in thread

From: Josh Berkus @ 2002-01-26 00:27 UTC (permalink / raw)
  To: jm.poure@freesurf.fr, "Josh Berkus" <josh@agliodbs.com>; Francisco Reyes <lists@natserv.com>; +Cc: pgsql-php

Jean-Michael,

> Are you connecting from Netscape? I noticed long queries timeouted in
>  
> Netscape. This is because Nescape waits for the entire content of a
>  page 
> before displaying it.

Yes, and several other Mozilla derivatives.  For example, Konqueror has
 a .conf variable that defaults its timeout at 60 seconds.  While one
 can adjust this, you can't count on hundreds of users doing the same.

> PostgreSQL can run very fast when well-optimized. You probably need
>  to 
> optimize the source code of queries themselves, probably writing some
>  good 
> PL/pgSQL server side code. If you need a GUI to do so, download and
>  install 
> Dave Page's excellent tool http://pgadmin.postgresql.org

Nope don't need a GUI.  The query in question was actually a procedure
 that evaluated 700-1500 candidates for jobs based on between 14 and 90
 different scalable criteria and then ranked them by suitability.  So
 there was a good reason for it to run for a long time ... that, and
 the DB server's damn Celeron chip which I eventually replaced.

Thanks for the general advice.  I'm sure the other list members will
 benefit from it.

-Josh

______AGLIO DATABASE SOLUTIONS___________________________
                                       Josh Berkus
  Complete information technology      josh@agliodbs.com
   and data management solutions       (415) 565-7293
  for law firms, small businesses        fax 621-2533
    and non-profit organizations.      San Francisco



^ permalink  raw  reply  [nested|flat] 16+ messages in thread


end of thread, other threads:[~2002-01-26 00:27 UTC | newest]

Thread overview: 16+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2002-01-19 16:48 ` Francisco Reyes <lists@natserv.com>
2002-01-19 17:32   ` Chadwick Rolfs <cmr@shell.gis.net>
2002-01-20 00:10     ` Mitch Vincent <mitch@doot.org>
2002-01-20 02:59       ` Cornelia Boenigk <poppcorn@cornelia-boenigk.de>
2002-01-20 14:56         ` jeff fitzmyers <fitz22@earthlink.net>
2002-01-21 17:02       ` Chadwick Rolfs <cmr@shell.gis.net>
2002-01-22 01:30         ` Christopher Kings-Lynne <chriskl@familyhealth.com.au>
2002-01-20 15:56     ` Francisco Reyes <lists@natserv.com>
2002-01-21 10:21       ` Marco Colombo <marco@esi.it>
2002-01-19 17:42 ` Frank Bax <fbax@sympatico.ca>
2002-01-20 15:50   ` Francisco Reyes <lists@natserv.com>
2002-01-22 15:52 Re: Long running queries and timeouts Josh Berkus <josh@agliodbs.com>
2002-01-23 03:54 ` Francisco Reyes <lists@natserv.com>
2002-01-23 20:39   ` postgresql@finner.de
2002-01-23 13:58 ` Jean-Michel POURE <jm.poure@freesurf.fr>
2002-01-26 00:27   ` Josh Berkus <josh@agliodbs.com>

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