Air Cleaner User Manual

The easiest way to capture stdout and stderr is to append this string to the end of
your backup command:
> /tmp/backup.log 2>&1
By adding this string instead, you will generate a unique log name based on the
date and time with each backup:> /tmp/backup.log$(date "+%m%d%H%M").log 2>&1
Unattended Backups Running in the Background
By default, if the backup command process is run through cron, nohup, or the “at”
command and there is no controlling tty because the backup is running in the
background, the backup process will terminate if the device (tape or virtual device)
is:
v Not Ready
v Contains write protected media
v Contains no media at all
If you would like the process to wait indefinitely to be resolved, rather than to
terminate, you should add the -A flag the: sysback, mkvgback, mklvback,
mkjfsback, and mkdirback, commands.
If you also use the -c /dev/termdev flag to the backup command, a message will be
displayed on the terminal when one of these conditions occur. If you do not use
the -c /dev/termdev flag, and you have used the -A flag, SysBack will check to see
if there is a controlling tty, and if there is not, it will use -c /dev/console.
When there is a controlling tty for the backup process, for example the backup is
run in the foreground, the backup process will wait indefinitely if any of the
aforementioned conditions are encountered. Once the condition is rectified, the
process will continue.
This functionality only applies for the first tape in a backup (i.e.: when the back
starts). If a backup spans to multiple tapes, and any subsequent tape meets any of
the described criteria, then the backup process will wait until such time as this
condition is corrected. Currently, the underlying commands will not display a
message indicating that this condition has been encountered and the backup
process may appear hung. This is a known limitation.
4-32 IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide