Hello,
sorry if this question was already asked - didn't find a good resource for explanation.
in short:
If i have 2 classes with oneToMany<->ManyToOne mapping (i.e. a Customer has many invoices):
For every generated invoice, there will be a revision of customer.
This will lead to excessive storage consumption (if you think about customers that get invoices every day without any data change)
Could anyone clarify which function realldy needs this revision?
I am interested in:
- customers state at invoice generation (for invoice address and so on) -> getRevisionFor(invoice.date) -> get customer for this revision
I am not interested in:
- At revision 8137 the customer object got a new invoice in its invoice collection (can get this through invoice date)
I hope the example clearly states my needs and you could give me a hint to avoid 50 million customer audition rows for 10000 customers without any customerdata change
Many thanks,
Tom
If you don't want to generate revisions for customers, if an element is added/removed to a relation's collection, just set the
org.hibernate.envers.revision_on_collection_change
property to false.
Adam