Disk Copying/Cloningwith LiLo errors
 

We have a Cisco CE-507 Content engine that uses Linux as the OS installed on an 18.2 GB drive. We tried to clone the drive for replacement. Now when we boot on the backup drive it freezes at LiLo. The primary drive still boots just fine. One idea was the drive we copied from (Disk0) would be different for (Disk1). But that makes no sense since we are doing a full disk copy, the files should be the exact same. We used Norton's Ghost 7.5 and 8 without success. Norton's was the only program we found saw the drives properly. We are doing this in a MCS-7820 (which is a Compaq Prosignia 720 server essentially). Is there anything we need to know beforehand when cloning a Linux partition/disk?

Always keep a rescue disk handy - you will in most cases have to reinstall the bootloader because of the changed disk geometry. With Lilo, it's as easy as running /sbin/lilo.


Håkan

Thats the thing that has us stuck-disk geometry. The disks are the exact same. The CE-507 comes with 2 18.2 GBs. The second was not used, so we want to make a quick backup of that should the first ever fail. As I understand, the geometry should not have changed when doing a clone of 2 drives of the same geometry correct?

Are you connecting the drives identically? I mean, do you disconnect the former primary disk altogether and connect the backup disk as the primary master? Also check that the BIOS uses the same access mode (LBA32/48?) for both disks.


Håkan

Since these disks are SCA SCSI I am using 2 SCA to 68Pin adapters. The primary disk which I want cloned is set to SCSI ID 0, while the second disk which I want to clone to is set to SCSI ID 2. Would it change if I used Ghost multicast and saved the copied disk to a network storage, then reconnect the second disk in place of the original and move the partitions back on? Thing is, someone here mentioned that LiLo directs boot in regards to disk drive, which is fine, but since we are just "copying" and not modifying I do not see why this does not work. If the original drive's LiLo referenced Disk0 and we put it in as Disk1 or 2 I can understand.

One thing, the second drive we have originally was clean in that it just came back from IBM via an RMA (which is why we want to protect the primary drive. The second drive failure caused no harm, but if the primary fails then we have a problem and a time consumeing task re-installing the OS)