Hi,

I have entity beans which act purely in a stateless service fashion -
essentially executing SQL and performing business logic and then
returning business objects.

I establish the JDBC connection manually and all the beans *seem* to be
using the same connection since it is accessed statically.  I dont know
whether that creates a problem if the beans are all their own threads..
anyway.

I am all for some pooling - I vaguely remember someone suggesting that
standalone JDBC can easily pool for you - without a container or
anything else.

If so I am not sure what I would get from an EJB datasource beyond some
transaction management possibilities down the road if I cant architect
around it.

So my question is a) can I pool with POJO and b) is it far better to
have the container do it?

Note I generate and execute my own SQL - no real entity beans here.

thanks

Tim