This is with regards to what I would term a fairly fundamental design fault, in Backup Exec (BEWS), which has existed for several versions. It specifically applies to single tape backup drives (not autoloaders) for backups run at small/remote sites, in an unattended manner.
Scenario:
A backup tape is inserted into the tape unit. For whatever reason, the tape is associated with the wrong media set and the overwrite protection period (OPP) or append protection period setting prevents the job from progressing. At this point the job enters a waiting state. While the job is in this state BEWS creates a media removal alert. What it also does at this point, and this is the point where the logic breaks down in BEWS, is to eject the tape, thereby making it impossible to rectify the issue and necessitating a visit to site to push the tape back into the drive!
What should happen is as follows:
The incompatible OPP should cause the job to enter its waiting state and BEWS should still create the media removal alert. However, the tape should not be ejected. The media removal alert will alert the operator to check the backup job and media properties without any harm coming to the data on the tape. The operator then has an opportunity to decide the next course of action. If it is to stop the backup job, then the operator hits the "Acknowledge" button on the media removal alert and, ONLY AT THIS POINT, BEWS ejects the tape. However, if the operator realises the issue with the media set on the current tape they can associate the job with the correct media set. Once they have done this they will be able to hit the "Cancel" button on the remove media alert and the tape will NOT be ejected. The backup job will then leave its waiting state, because the alert has been cleared and the job will find that the media is now in a backup set that will allow the job to continue.