Frequent log switches during sqlldr batch jobs

Hi,

we are running 9.2.0.6 on Solaris. Each night, a bunch of sqlldr batch jobs are run and the redo log switches per hour increase dramatically: 100 and more per hour, while during normal production hours we're at a maximum of about 25 (usually much lower).

However, I don't see any "Checkpoint not complete" messages in our alert file. Does it make sense to increase the redo log file size (The amount of data to be written would be same, wouldn't it)? What other possibilities do I have to check if there are performance problems when switching log files so often?

Thorsten Received on Wed Feb 02 2005 - 05:56:38 CST