I took a full disk image of a 4TB virtual disk (RAID50+2 on ServeRAID-M5014) from IBM X3650 M3 server.
The source system runs Debian 7.1 and uses LVM.
For the image I used a 6TB HGST AF disk in an Anker docking station connected via USB.
Then I attempted a restore to a 4TB virtual disk (slightly higher size than the source, it's RAID10+2 on PERC H730P Mini) on Dell PowerEdge R730.
The restore took 2 x 16 hrs rounds (never seen 2 long runs before) and on the second run it was explicitly prompting me for restoring every single LVM volume.
For backup and restore I used the latest Clonezilla 2.5.2-31-amd64.
The only difference being for backup I used bootable USB and for restore a live CD as the server failed to boot of the USB for some reason.
I didn't see any errors in the restore process until the very end: "Failed to install grub".
Some output generated from Debian 9 live CD running on the unbootable target Dell server:
fdisk -l
Disk /dev/sda: 3.8 TiB, 4196848631808 bytes, 8196969984 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: gpt
Disk identifier: 30A7726B-5F2A-4A55-B2CD-4D97433914AB
Device Start End Sectors Size Type
/dev/sda1 2048 4095 2048 1M BIOS boot
/dev/sda2 4096 249855 245760 120M Microsoft basic data
/dev/sda3 249856 7796865023 7796615168 3.6T Linux LVM
Where "Microsoft basic data" is coming from as both the source system and Clonezilla are Linux Debian based?
Is it something Clonezilla created trying to make the disk bootable?
(parted) print /dev/sda
Warning: Not all of the space available to /dev/sda appears to be used, you can fix the GPT to use all of the space (an extra 400102912 blocks) or continue with the current setting?
parted: invalid token: /dev/sda
Fix/Ignore? I
Model: DELL PERC H730P Mini (scsi)
Disk /dev/sda: 4197GB
Sector size (logical/physical): 512B/512B
Partition Table: gpt
Disk Flags:
Number Start End Size File system Name Flags
1 1049kB 2097kB 1049kB grub bios_grub
2 2097kB 128MB 126MB ext3 boot msftdata
3 128MB 3992GB 3992GB lvm lvm
Still not sure what the best way forward is...
Would it be of any benefit trying to edit volumes of the LVM partition (with LVM tools as direct mounting is not possible)?
Or maybe rather attempting grub2 install / repair?
Last edit: Adam Weremczuk 2017-10-12
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
First, backup important data before you use Clonezilla. Just in case.
Second, RAID is complicated, so Clonezilla does not fully support it. It might work, but might fail.
So which version of Clonezilla live did you use? Did you try the latest stable or even testing Clonezilla live? http://clonezilla.org/downloads.php
Steven
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I used the latest Clonezilla 2.5.2-31-amd64 live USB for backup and 2.5.2-31-amd64 live CD for restore.
I'm not too concerned about the data at this stage yet.
The destination server B is not needed for anything else and the source server A hasn't been overwritten yet (more details about my goals down below).
The goal of this exercise was to verify if the whole disk image taken by Clonezilla can be restored.
It doesn't look like it can be restored to a different RAID-ed hardware.
The more important question to me is:
What are my success chances of restoring back to the same (source) hardware?
Since I can't really prove I can restore this, it will remain a gamble.
I used Clonezilla in the past with great success for cloning single drives with physical partitions but never for RAID-ed servers with LVM.
So another question is if Clonezilla is the best choice for what I'm trying to achieve in a fairly automated and hassle free fashion?
Let me point out my (fairly complicated) scenario for clarity:
Back up server A to a single image.
Restore that image to server B (different hardware) to verify if restore works.
Backup server C to a single image.
This server has identical hardware as server A - chassis, disks, RAID card and config, firmware etc.
It has much more memory, and much lower "milage".
Restore image of server C taken in point 3 to server A.
Data destruction occurs here, the only backup of server A is now the image taken in point 1.
Restore image of A from point 1 to server C.
The ultimate goal is to hardware swap servers A and C.
Data of server C is much more important than A.
All servers use RAID and LVM.
Any advise would be appreciated.
Thanks
Adam
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
"What are my success chances of restoring back to the same (source) hardware?" -> As I have mentioned, RAID is complicated., and Clonezilla does not fully support it. Therefore actually I do not have any idea about the chance. You will only know after you test it.
Steven
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
That should make no difference since Linux can read and write ext4 or ntfs file system. Hence using either one of them as the image repository should be OK.
Steven
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello,
I took a full disk image of a 4TB virtual disk (RAID50+2 on ServeRAID-M5014) from IBM X3650 M3 server.
The source system runs Debian 7.1 and uses LVM.
For the image I used a 6TB HGST AF disk in an Anker docking station connected via USB.
Then I attempted a restore to a 4TB virtual disk (slightly higher size than the source, it's RAID10+2 on PERC H730P Mini) on Dell PowerEdge R730.
The restore took 2 x 16 hrs rounds (never seen 2 long runs before) and on the second run it was explicitly prompting me for restoring every single LVM volume.
For backup and restore I used the latest Clonezilla 2.5.2-31-amd64.
The only difference being for backup I used bootable USB and for restore a live CD as the server failed to boot of the USB for some reason.
I didn't see any errors in the restore process until the very end: "Failed to install grub".
A final screen shot below:
https://drive.google.com/file/d/0B2GgcGgLUlJgWFFGVktnZWlHZDA/view?usp=sharing
As a result the destination server fails to even start booting from the local disk.
Does anybody know what went wrong and how to fix it?
Is performing the same full restore to the original hardware more likely to work?
Thanks
Adam
Last edit: Adam Weremczuk 2017-10-12
Some output generated from Debian 9 live CD running on the unbootable target Dell server:
Where "Microsoft basic data" is coming from as both the source system and Clonezilla are Linux Debian based?
Is it something Clonezilla created trying to make the disk bootable?
Still not sure what the best way forward is...
Would it be of any benefit trying to edit volumes of the LVM partition (with LVM tools as direct mounting is not possible)?
Or maybe rather attempting grub2 install / repair?
Last edit: Adam Weremczuk 2017-10-12
First, backup important data before you use Clonezilla. Just in case.
Second, RAID is complicated, so Clonezilla does not fully support it. It might work, but might fail.
So which version of Clonezilla live did you use? Did you try the latest stable or even testing Clonezilla live?
http://clonezilla.org/downloads.php
Steven
Hi Steven,
I used the latest Clonezilla 2.5.2-31-amd64 live USB for backup and 2.5.2-31-amd64 live CD for restore.
I'm not too concerned about the data at this stage yet.
The destination server B is not needed for anything else and the source server A hasn't been overwritten yet (more details about my goals down below).
The goal of this exercise was to verify if the whole disk image taken by Clonezilla can be restored.
It doesn't look like it can be restored to a different RAID-ed hardware.
The more important question to me is:
What are my success chances of restoring back to the same (source) hardware?
Since I can't really prove I can restore this, it will remain a gamble.
I used Clonezilla in the past with great success for cloning single drives with physical partitions but never for RAID-ed servers with LVM.
So another question is if Clonezilla is the best choice for what I'm trying to achieve in a fairly automated and hassle free fashion?
Let me point out my (fairly complicated) scenario for clarity:
Back up server A to a single image.
Restore that image to server B (different hardware) to verify if restore works.
Backup server C to a single image.
This server has identical hardware as server A - chassis, disks, RAID card and config, firmware etc.
It has much more memory, and much lower "milage".
Restore image of server C taken in point 3 to server A.
Data destruction occurs here, the only backup of server A is now the image taken in point 1.
Restore image of A from point 1 to server C.
The ultimate goal is to hardware swap servers A and C.
Data of server C is much more important than A.
All servers use RAID and LVM.
Any advise would be appreciated.
Thanks
Adam
"What are my success chances of restoring back to the same (source) hardware?" -> As I have mentioned, RAID is complicated., and Clonezilla does not fully support it. Therefore actually I do not have any idea about the chance. You will only know after you test it.
Steven
Could the filesystem on the drive used for storing images (e.g. NTFS vs EXT4) potentially make any difference?
In my case it's NTFS.
That should make no difference since Linux can read and write ext4 or ntfs file system. Hence using either one of them as the image repository should be OK.
Steven