[Pgpool-general] pgpool problems
Tatsuo Ishii
t-ishii at sra.co.jp
Thu Feb 24 08:02:08 GMT 2005
> Hi Tatsuo,
>
> greetings to Japan!
greetings to Denmark!
> > pg_pconnect will eat up all pgpool child processes(number of them is
> > 64, in your case). If you want to keep using pg_pconnect, you need to
> > increase num_init_chidren until it exceeds number of possible Apache
> > child processes (probably this is defined by MaxClients directive in
> > httpd.conf).
> >
> > Or changing all pg_pconnect calls to pg_connect should solve your
> > problem.
>
> I tried with pg_connect and pg_pconnect, but maybe I should restart all
> webservers afterwards. I'll test this again.
>
> Do I assume correctly that if I have a large percentage of "idle"
> postgresql processes I may lower num_init_children and raise max_pool?
If your web application connect to PostgreSQL as all same user, say
"nobody", raising max_pool will not benefit at all. Actually in this
case max_pool = 1 is sufficiant. However since max_pool slot eat very
little memory, you might want to leave as it is.
Changing num_init_children will affect the total throughput. In
theory, lowering num_init_children will decrease the load of
PostgreSQL but increase the chance for pg_connect waiting for next
available pgpool's child process. Increasing num_init_children will
have opposite effect. So there must be a sweet spot for
num_init_children. I suggest you play around several num_init_children
settings seeking for the sweet spot.
--
Tatsuo Ishii
More information about the Pgpool-general
mailing list