From:Steve Adams
Date:17-Oct-2000 07:58
Subject:   redo log space requests

Run the following statement repeatedly, say every second. 

    select
      l.sequence#,
      s.value
    from
      v$log  l,
      v$sysstat  s
    where
      l.status = 'CURRENT' and
      s.name = 'redo log space requests'
    /
You will probably see that the redo log space requests correlate very strongly with log switches. If so, it is the speed of log switches that you should be tuning, rather than the size of the log buffer.

We tried that. We increased it from 1M to 2M and the messages kept coming. How can I see if the requests are coming at that time?

The Ixora web site contains plenty of information on this topic. Nevertheless, the bottom line is that you should increase the size of your log buffer if and only if the redo log space requests are happening at times other than immediately after a log switch.

We get the following message many times a day: 
    X < 4000 redo entries per redo log space requests
We are running SAP on Oracle 8.0.6. Is this due to the redo log size (20M) or the log_buffer parameter (1M)?