I, too, was having difficulties with a distributed transaction. Does anyone know Oracle's rationale for not allowing a remote proc to not perform commits? Is it because it was too difficult for them to have the commit actions synchronized if they could be performed in more than one location? Operators are standing by waiting for your answers...
If you think about it, remote commits don't make sense. The transaction is initiated on one side, why should the other side be allowed to take over control of that transaction?
Bookmarks