Hi all, I'm just getting started with hibernate, and I am no DB expert...

Anyways the basic question is: Given A Query may return say 1 million rows, retrieving all 1 million records into a list..is obviously going to be memory issue. so..should the query result in in that many matching records, how can I specify a max return size?, and then or course continue the query later form the last record previous retrieved.

i.e. Doing:
Query q = session.createQuery("from BigMoFoTable");
List results = q.list();

could potentialy result in a list getting created with all the records?
(I'm assuming the list just wraps the JDBC ResultSet).

is there anyway to say do something along the lines of...

Query q = session.createQuery("from BigMoFoTable");
List results = q.list(MAX_SIZE);
.. DO WORK..
// Get the next 'batch' OF RESULTS
q.list(LAST_RESULT_SET_INDEX, MAX_SIZE);


THANX FOR ANY advice/tips, this noob will greatly appreciate it.

Chap 7, has the answer.

why don't you use iterators? by the way you can set first result index and max number of results on query and criteria interfaces.

actually have indeed started using iterators instead, the only downside it would be nice to the pass back the 'number' of potential results, when getting the iterator, I suppose I could do a getCount() type query if this was needed.

As of right now my plan is:

Use Lists for 'base data' type tables (and splice them into maps/lists) for quick jsp access.
Use iterators where the result sizes are unpredictable, and could be large..

Still need to read up hibernates caching stratgies, but I suspect it can cache 'iterators' accross sessions....

ps. anyone know of more upto date hibernate books than Hibernate in Action...(which Orielly had Hibernate book out)

I strongly recommend that you download the hibernate source and hit breakpoints to see what is going on behind the scenes. There is no better learning tool that the source itself. Take a look at Loader.bindLimitParameters().

What you'll discover is that the limit/offset functionality is highly dependent upon your RDBMS, and you'll need to watchout for performance issues surrounding the availablity of these features and how your JDBC driver might compensate for them. For example, check out MS SQL Server, which has TOP, but no offset. This results in some client-side or server-side cursors that might not be too efficient, depending on the situation.