Just wanted to address some of the questions. The os is a current version of Red Hat. X11 is not installed, so no gui. For a program like this I usually run it from cron on a detached
. I'm reading the output becuase I had code which was calculating and printing the bytes/second read and the wall time the dump has been running. I decided to use a freeware
instead, so I can probably remove that block. The machine in question has 8 cpus and plenty of IO bandwidth so I believe dumping two different databases at once would speed up the backup process (which involves duping around 3 gb of data).