Purpose of rfs

rfs stands for "replication of file systems".

It is a shell script for creating and updating a local spare system disk. The main goal is to recover a working system after a crash quickly. In this case, "quickly" means the time it takes to reboot the machine. The very first version of rfs was clonedd.sh, a shell script written by Eric Gerbier.

rfs works on Gnu/Linux platforms and supports ext2/ext3 file systems. Targets are workstations as well as servers : your system remains fully-working while the backup is processed.

Terms of use

rfs is free and licensed under the General Public License : you can use it, modify it at your convenience. If you plan to embed it in a (commercial) package, remember this package has to conform with the General Public License. See file COPYING for details.





The current release is rfs-2.13-4.tar.bz2


tar -xjf rfs-2.13-4.tar.bz2
cd rfs-2.13-4
su -
make install
To build a rpm : rpm -tb rfs-2.13-4.tar.bz2 See RPM-HOWTO to have a further look on building a rpm.


To clone __c /dev/hda __ on __c /dev/hdb __ , become root and type : rfs /dev/hda /dev/hdb If your system disk crashes, put the target disk in place of the source one, and reboot.


Bug report

  1. use the last available version of rfs.
  2. send a log to Eric from a command like : rfs -f -v -D /dev/hda /dev/hdb 2>&1 | tee /tmp/rfs.log


To contact us, please mention "rfs" in the subject of your email. Copyright (C) 2004 2005 2006 2007 Nicolas Jouanne