We were forced to do a shutdown abort, then started the instance up
normal, but a shutdown immediate right after that hangs every time.  The
alert.log does not show any errors or trace files.  Any ideas?