Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

What is the approach generally used when you want to have backup physical servers?

Currently I have a Linux server running a database, a samba share, a webapp and some scripts; and a Windows Server, running some third-party software. What I would like was to be able to have a ready backup server to enter in production in case of failure, but how to keep them up-to-date? I've seen some expensive solutions for Windows; for Linux I've wondered if I really have to build an array of scripts.

share|improve this question
up vote 0 down vote accepted

You are talking about a general cold-standby solution. This is a very broad topic with several possible solutions. The following are just a few of them.

For the database I would really recommend looking into a standby server. MySQL got a good slave-replication, other databases got that too.

For files the easiest thing would be using rsync in daemon mode. You could also look into block level replication tools like DRBD or distributed file systems, but those might hurt performance.

You can also create images using a combination of LVM and snapshots. Or you could use a tool like Image for Linux(which is commercial software, but pretty cheap for what it can do).

For configurations you can either use some sync-scripts or rsync or even some configuration management like Puppet.

share|improve this answer
Thanks; would config solution with Puppet work on Windows? this is not my primary job, so I've tried to research some imaging solutions that would make my life easier. Turns out there's expensive Windows solutions, but no Linux solution other than open source software which "only" does cold image backups. – ftkg Jun 24 '13 at 22:14
If I remember correctly there was Windows support for Puppet, but I don't have any Windows servers in my network, so I can't tell how good that works. Also are you sure you want to take images? In most cases replication is way more flexible than just bare images(and reduces load and bandwith). – mkzero Jun 24 '13 at 22:18
I don't know really, was just trying to find some practical solution. Building a mirror server and learning Puppet might be too time-consuming for me. – ftkg Jun 24 '13 at 22:27
It all depends on the allowed downtime between server failures. The less downtime, the more time it takes. If you can have several hours and more, images might be just okay(added some stuff to my answer to tackle that). But I think you would have to research bit more yourself before continuing. – mkzero Jun 24 '13 at 22:33
My downtime requirements is not that critical for the Windows server; if I can restore a image in 2-3 hours it should be fine. Didn't know that tool, thanks. They have a bare-metal solution, will try to check it out. – ftkg Jun 24 '13 at 22:45

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.