Thanks for all of the help last time . I installed Backupninja easily. After getting it running I realized that it didn't quite do what I wanted. I was looking for a package that could be installed in the server with little need for additional programs to be installed on each machine. So, after some searching, I switched to Backuppc. Now Backuppc will do what I want but it needs to run first. To cut to the chase I have (I think) the config.pl file set to backup /home, /etc, and /var with a couple of exclusions. As the debian readme file suggested, I pasted the supplied script into the Apache2 config file. This seems to be working OK. I then moved the /var/lib/backuppc files to my spare hard drive '/backupdisk' and did a soft link from /var/lib to /backupdisk/backuppc. I then did the following
root/.../etc# service backuppc start
Starting backuppc...2012-02-09 17:17:05 Can't create a test hardlink between a file in /backupdisk/pc and /backupdisk/cpool. Either these are different file systems, or this file system doesn't support hardlinks, or these directories don't exist, or there is a permissions problem, or the file system is out of inodes or full. Use df, df -i, and ls -ld to check each of these possibilities. Quitting...
I searched for an answer and found dozens, all different and mostly old. The either made no sense or didn't work. Specifically- all of my disks on all of my machines are ext3 file systems. Changing the permissions to 755 on all of the files didn't work.
The path's involved look like:
/var/lib/backuppc -> /backupdisk/backuppc
/backupdisk/backuppc/ cpool, pc etc.
I am running Debian Squeeze on all systems
I am using rsync and have ssh installed on all systems.
All of the systems communicate with each other over ssh
My backup server is running two GB size disks. Backuppc is running on one disk and using the other for the backup data.
Any ideas of how to fix this problem will be sincerely appreciated.
PS: For the average user the documentation for this package sucks. The detail included may be exactly what the developer needs but is really rough sledding for the average user. This seems to be a very common problem in the Linux world and is the main reason that Linux is still relegated to the number three spot way behind Windows and Mackintosh. I spent 40 year in industry and was a tech writer for many of those years. My last 5 years was as a developer of medium size database systems for use by real dummies. So I speak with some authority. Every program that is for general use must have one of two things. Either a GUI that can be used without any special training by your wife/girlfriend/secretary without asking any questions or documentation that passes the same test. This is the criteria that I always used and it works. I would love to rewrite some of this stuff but am not nearly conversant enough with the nitty gritty programming details.