Why does EclipseLink think my database schema doesn't exist?

Question

I'm working on a web project using EJB 3.0, and whenever EclipseLink tries to interact with the database, it says that the schema I'm using doesn't exist (which it does).

I get a massive, unhelpful stack trace from GlassFish 2.1, which begins with:

EclipseLink, version: Eclipse Persistence Services - 1.1.0.r3639-SNAPSHOT
file:/C:/Documents%20and%20Settings/nsteinberg/.personalDomain/personalDomain/applications/j2ee-apps/ScienceEar/lib/ScienceJpa-1.0-SNAPSHOT.jar-SciencePU login successful
Local Exception Stack: 
Exception [EclipseLink-4002] (Eclipse Persistence Services - 1.1.0.r3639-SNAPSHOT): org.eclipse.persistence.exceptions.DatabaseException
Internal Exception: java.sql.SQLSyntaxErrorException: Schema 'BEAMLINE' does not exist
Error Code: -1
Call: SELECT exp_id, fac_family, public_viewable, group_name, created_ts, status, exp_num, date_received, exp_type, title, updated_ts, text_only_title, experiment_url, proposed_eec, start_of_prep FROM beamline.eec_exp_toc
Query: ReadAllQuery(name="Experiment.findAll" referenceClass=Experiment sql="SELECT exp_id, fac_family, public_viewable, group_name, created_ts, status, exp_num, date_received, exp_type, title, updated_ts, text_only_title, experiment_url, proposed_eec, start_of_prep FROM beamline.eec_exp_toc")

I could post more of the stack trace, but it's really boring. Any ideas for why EclipseLink can't see the schema?

More info: "beamline" is the only schema I use. The SQL in the stack trace is the stuff that was generated by a NamedQuery, the first interaction with this database (and the first JPA interaction at all) in my program.

Answer 1

Got it - I'm connecting directly to the database, when I should be using GlassFish's connection pool.