Too Many Log file sync events reported in AWR Reports

One possibility to fix the error:

A database might experience sudden spikes in the average "log file sync"
wait time with ARCHIVE_LAG_TARGET set.
 
Rediscovery Notes:
Look for:
 - ARCHIVE_LAG_TARGET is set to some non-zero value
 - performance monitoring shows occasional spikes in the
   "log file sync" wait times (eg spikes are seen on a 
   graph plotting the time of day in seconds along the 
   x-axis, against the average "log file sync" wait time 
   on the y-axis)
 - routines on LGWR's call stack during the spikes show:
    -> ksbrdp -> ksbabs -> ksbcti -> kcrfwl ->
      -> kccocx -> kcc_get_enqueue ->
 - the contents of V$LOCK during the spikes show LGWR
   waiting for the CF enqueue in SSX-mode (REQUEST=5):
     select vs.program,lmode,request
     from v$lock vl, v$session vs
     where vs.sid=vl.sid and
           vl.type='CF' and vl.id1=0 and vl.id2=0;
 - setting ARCHIVE_LAG_TARGET to zero resolves the issue 
   (this can be done dynamically)
 
Workaround
 Set archive_lag_target to zero.