DRCP is really a big improvement. But for me it has a great holdback:
the pool controlled in the server side isn't defined for a specific
schema, so if we have an application server serving different
connections demanding different credentials certainly the performance
will be worse. Another big problem is that one application hosted in the
web server can lock the pool completely is the demand for this
application improve.
In other hands if we work with traditional pooling, configured in the
client side, we should define one pool for each aplication with his own
credentials. The pool will never blocked by one system and for hosts
with several credentials the performance will be better. The performance
and availability will be insured.
Of corse both are good solutions, each one for your specific purpose.
You have to make a change to the php code (the tns definition) to use
pooling - it's not going to use pooling just because you enable it.
--
Postgresql & php tutorials
http://www.designmagick.com/
--
PHP Database Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php
