Re: [HELP]: Large cursor updates

In article 16381_at_KYnug.org, neil_at_kynug.org (Neil Greene) writes:
>I have a cursor that scans a table of approximatly 10,000 records and does
>various inserts to a table based on what it finds. When the same cursor
>is run on a select statement that finds a smaller selection of records,
>say 100-1,000, everything seems to complete quickly and properly.
>However, cursors that have a larger found set appear to hang.
>

I seem to have the same problem using PL/SQL (2.0.14.0.1). I have a table of ~500,000 records which I am trying to copy into another table doing some data conversion on the way and inserting summaries into some other table, the work is commited after every 1000 records fetched. Because my PL/SQL cursor has to scan the source table in certain order I have created appropriate indexes on that table. When I test the program on a smaller subset of data (few thousand records) everything goes ok. When I run it on the full set it runs for days without producing any visible results. Splitting the task into 100 smaller tasks each processing only part of the table is not practical. Any suggestions?

P.S. HAPPY NEW YEAR TO ALL NETTERS!!!

Miroslaw Ryba                     _--_|\        ryba_at_ml.csiro.au 
CSIRO Marine Laboratories        /      \       tel +61 02 325 336 
PO Box 1538   Hobart 7001        \_.--._/       fax +61 02 325 000
Australia                              v