Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Usenet -> c.d.o.server -> Re: trying to understand transaction control in pl/sql
Hi,
I need to add a note to my previous post - it rolls back until it finds a handler for the error - all the way to the sqlplus prompt or equivalent if necessary - BUT IF IT FINDS ONE it only rolls back the one statement that errored - so it must not do the actual rollback until it determines it's necessary.
I suppose it marks wherever it stores the new data with all the intermediate savepoints so it can do this partial write/partial rollback if an error occurs which is eventually trapped.
Thanks,
Ken
PS it makes sense to talk of rolling back a statement that causes an error because for example there might be a trigger executing after the insert that causes the error. That make sense? Received on Mon Nov 01 2004 - 07:03:19 CST
![]() |
![]() |