This is a good technique, but let me clarify the problem.  When a drive is
polled, the busy flag
gets turned on during the active part of the polling.  This is so that you
cannot delete the drive
until the poll gets back, which could abend the server.   The polling is every
30 seconds, and this
is only for a few seconds of that time.  (the polled flag stays on
constantly.)  It may just be
coincidence that you saw the busy flag set to 1 each time you checked.

If the drive was polled, then there was some problem getting to it.   I would
have thought that the
rmdev/mkdev would have done what the cfgmgr did as far as re-establishing the
connection
to the drive.   Thumping the lmcpd might have cleared any unsuccessful
mount/dismount to the
drive that would have caused the polling.

Craig Bell