generated SQL

[yanor]

has anyone noticed strange and inefficient SQLs genereated , like the one below? i haven't pinpoint exactly when the SQL is generated but i believe it only happens when i update a many-to-many relationship (i only have many-to-many uni-directional relationships so i don't know if this only happens in the uni-directional case). also, the number of "OR" conditionals depends on the number of objects in the table.

11:36:56,709 DEBUG [InvoiceEJB] Executing SQL: SELECT pk,name, product
FROM INVOICE WHERE (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)
OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?) OR (pk=?)

[pazu]

This is JBoss trying to fill the entity pool with not-yet-loaded entities.

[sonwh98]

Can you explain this further? I am not familiar with the internals of Jboss, but it seems executing such an SQL query is not efficient.

[dsundstrom]

This is block loading of entities, and it helps to avoid many sequential loads of single rows. It is discussed in the JBossCMP documentation.