On 21 Nov 2012 14:45, Dan Kimmel wrote:
Removing a member (RMVM) doesn't necessarily free the space on the
FWiW: Unless a defect [or a usage issue for unpaired ALCOBJ\DLCOBJ
requests] prevents the database from destroying the dataspace of the
removed member, the RMVM will necessarily always "free the space" [i.e.
return the storage] for the member object and the data. The design
however is to be /tolerant/ of errors, so the RMVM affected by any
condition for which the data is still in use yet the *MEM object is
available to be destroyed, the only notification will be the joblog and
the VLIC logs; i.e. no exception message is signaled to the invoker of
the Remove Member request.