In order to fine tune the redo buffer cache, execute the following query continuously over a period of time. If the value is increasing steadly, it indicates redo buffer cache is undersized.
sys>select name,value from v$sysstat where name='redo buffer allocation retries';
If the value is increasing, increase the parameter value 'log_buffer'.
