Hi,
I'd to process data of a table in parallel using DBMS_PARALLEL_EXECUTE. I need to keep a single chunk rather small (about 100 rows) but this doesn't work.
I tried
dbms_parallel_execute.create_chunks_by_rowid(
task_name => g_task_name,
table_owner => user,
table_name => 'QUEUE',
by_row => TRUE,
chunk_size => 100 );
However, the chunk size was much larger, 1500 rows or more. I know that chunk size is just an approximation, but even when I tried setting the chunk size to a single row or block, the smallest chunk I could get was about 500 rows. I need small chunks, since I want to send the data later via HTTP and the requests should be rather small. I could split each chunk manually, but I would prefer to avoid this extra work. Any idea?