I have heard that there are domain controller and backup domain controler in MS windows system. My understanding was that if domain controller server fail to run due to hardware failures, then back up server acts like domain controller. I guess I do not have a good concept of this things.

However, I am sure about what I want to get. I have a server running Apache and Samba servers. I also installed many software. I spent much time to configure those. In this server, I wrote some scripts. These scripts are working fine. If this server stop working due to hardware failure such as hard disk problem, I probably need to get new hard disk and install and configure all again.

Therefore, I am considering to have backup server running with enough hard disk space. When the server stop working and not able to cure it quick, the backup server work just like original server. If this is possible, I can have time to repair the hardware problem. When the server recovered, I want the server works as a backup server automatically. So, I can have a server computer which works 24/7

Is this possible in Debian Linux?

Hi, I would suggest looking into mirroring, this should be an easy process and fairly easy to do with a cron job and rsync, I just did a quick google search and was able to find the following article that should be useful for mirroring two servers in a fairly basic method. This will say how to back-up/mirror the servers.

http://www.howtoforge.com/mirroring_with_rsync

As for the fail-over to the secondary server, this may need more work... unfortantly I don't have much I can offer in terms of advice on that front. I believe there are solutions that can be done with a DNS to point to multiple IPs or fail-over ips but I don't think it's perfect and I am not sure of how such is accompilised I am afraid. There are other methods of fail-over as well... but I have little knowledge of them

I would also point out that if your gunna go for such redundancies then you should have the two servers preferably in different locations/networks as well else you still have single failure points for both servers together.

P.S. I think this should be in the Linux - Server forum, this is just more then a bit above newbie level.