Prev: phpmailer/mysql/ timout issues
Next: Postgres - how to skip 'sequence' columns
From: Chris on 11 Apr 2010 19:26
David Murphy wrote:
> As you can see PHP claims it took 20 seconds for mysql->query() to return
> but mysql think is took around 1.0s
>
>
> This is from our application
> I enabled profile in mysql to determine why an update took 20seconds. As
> you can see MySQL reported no where near that amount of duration took
> place.
> Is there any way I can dig into php and determine why mysql client libs are
> so slow (this is not using mysqlnd but mysql-client-libs on CentOS using
> 5.3.2)

Is this a one-off thing or is it happening all the time?

If it's a one-off thing it could be a spurious result (maybe someone
else was doing a mysqldump when your query ran, the dump blocks your
query)..

What sort of mysql table is it? if it's innodb you can try it in a
transaction and roll it back:

begin;
update blah;
rollback;

see how long it takes.

if it takes a short time in the mysql client, then try it in a php
script from your other server.

--
Postgresql & php tutorials
http://www.designmagick.com/

 | 
Pages: 1
Prev: phpmailer/mysql/ timout issues
Next: Postgres - how to skip 'sequence' columns