How to create a full backup of a remote ubuntu 16.04 server?

Question

I just finished configuring a dedicated remote ubuntu server and want to create a backup with DD in order to be able to rebuild fast in case of hardware issues.

The system has 2 SSD drives with 500GB and a remote FTP server ready for backup. I created the following backups, but I am not sure if this was correctly done and would allow me to recover like intended. As it is a software raid, I assume it is enough to backup MBR and one SSD

Question: Is this backup procedure OK and sufficient to recover to the same server and what would be the savest way to rebuild from the backup files?

root@rescue ~ # lsblk
NAME                     MAJ:MIN RM   SIZE RO TYPE  MOUNTPOINT
sdb                        8:16   0 465.8G  0 disk  
|-sdb2                     8:18   0 465.3G  0 part  
| `-md1                    9:1    0 465.1G  0 raid1 
|   |-vg0-swap           253:1    0     4G  0 lvm   
|   |-vg0-pro2          253:6    0    24G  0 lvm   
|   |-vg0-pro1         253:4    0    30G  0 lvm   
|   |-vg0-tmp            253:2    0     5G  0 lvm   
|   |-vg0-root           253:0    0    10G  0 lvm   
|   |-vg0-staging_pro1 253:7    0    30G  0 lvm   
|   |-vg0-restore        253:5    0    30G  0 lvm   
|   `-vg0-home           253:3    0   140G  0 lvm   
`-sdb1                     8:17   0   512M  0 part  
  `-md0                    9:0    0 511.4M  0 raid1 
loop0                      7:0    0     2G  1 loop  
sda                        8:0    0 465.8G  0 disk  
|-sda2                     8:2    0 465.3G  0 part  
| `-md1                    9:1    0 465.1G  0 raid1 
|   |-vg0-swap           253:1    0     4G  0 lvm   
|   |-vg0-pro2          253:6    0    24G  0 lvm   
|   |-vg0-pro1         253:4    0    30G  0 lvm   
|   |-vg0-tmp            253:2    0     5G  0 lvm   
|   |-vg0-root           253:0    0    10G  0 lvm   
|   |-vg0-staging_pro1 253:7    0    30G  0 lvm   
|   |-vg0-restore        253:5    0    30G  0 lvm   
|   `-vg0-home           253:3    0   140G  0 lvm   
`-sda1                     8:1    0   512M  0 part  
  `-md0                    9:0    0 511.4M  0 raid1 

// MBR

dd if=/dev/sda bs=512 count=1 | gzip -fc | lftp userxx.your-backup.de -u uxx,pw -e "put /dev/stdin -o backup-MBR-$(date +%Y%m%d%H%M).gz; quit"

// SSDs

dd if=/dev/sda2 bs=2048 conv=noerror,sync | gzip -fc | lftp userxx.your-backup.de -u uxx,pw -e "put /dev/stdin -o backup-sda2-$(date +%Y%m%d%H%M).gz; quit"

// partition table

sfdisk /dev/sda –d | gzip -fc | lftp userxx.your-backup.de -u uxx,pw -e "put /dev/stdin -o backup-sda-part-table-$(date +%Y%m%d%H%M).gz; quit"

The backup files look OK:

ls -lh
-rw-r--r--   1        486 Feb 10 12:51 backup-MBR-201702101351.gz
-rw-r--r--   1        143 Feb 10 15:56 backup-sda-part-table-201702101656.gz
-rw-r--r--   1      86.7M Feb 10 15:54 backup-sda1-201702101654.gz
-rw-r--r--   1     182.3G Feb 10 15:42 backup-sda2-02-201702101440.gz

Answer 1

You can't create dd backups of a running system, as you most likely won't end up with a consistent image of your file system. To image something with dd it must not be mounted or mounted read-only.

Use one of the myriad other tools to do a proper backup.

If you dd from a rescue system instead of a live system, you could as well backup the whole of sda (or sdb) at once. Currently, you appear to be missing /dev/sda1, which quite likely is raided to be mounted as /boot, containing your kernel (which is essential, naturally). Also, if you have a MBR partition table, the partition table will be already included in the MBR (this is not true for GPT, of course).