How to copy a Linux installation

Using dd to make a 1:1 copy

Advantages

When dd is used to create the copy, it is copied at a sector by sector level. This means that in many circumstances, the new version will have it's boot sector set up and everthing. In situations where you are making a complete copy of the disk, even things like the IBM tools will be saved.

Disadvantages

Because dd copies sector by sector it copies everything, including sectors that aren't allocated. This makes it a longer process. To update the backup copy, it must perform the whole process again, rather than simply updating the things that have changed.

Case 1: The Linux installation is on a separate Harddisk

dd if=/dev/hd[a,b,c,..] of=/dev/hd[a,b,c,..] bs=2M

Part of Sourcedrive  : if=/dev/hd[a,b,c,..] the Letter "a" for the first Harddrive, b for the second, ....

Part of Destinationdrive  : of=/dev/hd[a,b,c,..] the Letter "a" for the first Harddrive, b for the second, ....

Case 2: The Linux installation is on a Partition

(e.g. hda1 is the Partition with the Linux installation and hdb1 is the Destinationdrive)

dd if=/dev/hda1 of=/dev/hdb1 bs=2M

Using tar to make a copy of the filesystem

Advantages

you're able to restore it to another filesystem