Maybe the problem is that you are doing too many log switches not that the alert log is filling up. If you use Unix you can probably create a shell script that runs in a cron to either delete the alert log of rename and compress. Oracle will recreate the file, empty and start filling it again. Depending on the amount of DML if you are doing more than 20-30 log switches an hour during peak utilization you may want to increase the size of the log files. You may also want to add additional log groups. There are scripts that give you information in waits on this site that might give you some insight. I'm not sure how you suppress the log switch message in the alert log or why you would want to. That is one way of looking at how busy your database is and that it is doing what it is supposed to be doing. There are many other ways of finding out if your database is working properly of course.