Re: COMMIT's within cursor for loops

2003-06-26 Thread Reginald . W . Bailey
I prefer to create an inner block using BEGIN and END inside the loop. This isolates the DML statements. The COMMIT is issued inside the BEGIN and END block. RWB Reginald W. Bailey IBM

RE: COMMIT's within cursor for loops

2003-06-26 Thread Steve McClure
I prefer to create an inner block using BEGIN and END inside the loop. This isolates the DML statements. The COMMIT is issued inside the BEGIN and END block. RWB That allows you to avoid invalidating the cursor established for update? I dont' see how the two are related. You just created an

RE: COMMIT's within cursor for loops

2003-06-26 Thread Ron Thomas
: [EMAIL PROTECTED]Subject: RE: COMMIT's within cursor for loops .com

RE: COMMIT's within cursor for loops

2003-06-26 Thread Jared . Still
to ORACLE-L To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] cc: Subject:RE: COMMIT's within cursor for loops I prefer to create an inner block using BEGIN and END inside the loop. This isolates the DML statements. The COMMIT is issued inside

RE: COMMIT's within cursor for loops

2003-06-25 Thread Goulet, Dick
You've got the most important point correct. When you issue that commit in the middle of a cursor all locks on the data tables are released. To be technically correct the cursor is also invalid and with any other database management system you'd have to close and reopen the cursor.