My memory was about an incident that occurred at an installation that
I worked at. Their DASD were 2314s (which were 9 drive units - 8 live
and one spare). The address of the drives were controlled by a round
plug that could be removed and placed in the address hole of another
drive. There were 2 of these units in the room addresses as 170 to
177 and 178 to 17F. The operator know that a job was going to start
which was going to need a certain disk volume so to save time he
placed it in the space drive of the 170-177 unit. When the job
started, the system unmounted 17A and asked for the volume to be
mounted there. To save time instead of spinning down the space where
he had mounted the volume and placing it in the drive currently known
as 17A (or the 178-17F Spare), he just moved the 17A plug into the
spare's address hole. Since there was no difference between a xx2 and
xxA plug (both are just indicate that they are the 3rd of 8 addresses
on the control unit - the 2314 control unit was just wired to respond
as 170-177 or 178-17F) suddenly there were 2 drives marked as 172.
You can imagine the crash that then occurred killing a long running
job using the real 172. The job had to be restarted from the
beginning after the volume was restored.
The operator was new there and was used to 3330s which only had 8
drives and no spares so this type of premount error could not occur
since the new volume would have either been placed in a drive whose
volume had been unloaded (and the system would have spotted it there
and selected it) or it would have done the unload to free up the
drive.
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN