Hi all!

I have a Debian stable server with two hdds in a md RAID which contains an encrypted ext-4 filesystem.

sda          8:0    0   2.7T  0 disk
├─sda1       8:1    0     1G  0 part                    
│ └─md0      9:0    0  1023M  0 raid1 /boot             
├─sda2       8:2    0   2.7T  0 part                    
│ └─md2      9:2    0   2.7T  0 raid1                   
│   └─mdcrypt                                           
│          253:0    0   2.7T  0 crypt /                 
└─sda3       8:3    0     1M  0 part
sdb          8:16   0   2.7T  0 disk                    
├─sdb1       8:17   0     1G  0 part
│ └─md0      9:0    0  1023M  0 raid1 /boot             
├─sdb2       8:18   0   2.7T  0 part
│ └─md2      9:2    0   2.7T  0 raid1                   
│   └─mdcrypt
│          253:0    0   2.7T  0 crypt /                 
└─sdb3       8:19   0     1M  0 part

I’d like to migrate that over to BTRFS to make use of deduplication and snapshots.

But I have no idea how to set it up since BTRFS has its own RAID-1 configuration. Should I rather use the existing MD array? Or should I take the drives out of the array, add encryption and then add the BTRFS RAID inside that?

Or should I do something else entirely?

  • Ooops
    link
    fedilink
    arrow-up
    4
    ·
    2 months ago

    Decryption isn’t a problem if you use the systemd hooks when creating your initrams. They try to decrypt every given luks volume with the first key provided and only ask for additional keys if that fails.

    I have 3 disks in a btrfs raid setup, 4 partitions (1 for the raid setup on each, plus a swap partition on the biggest disk), all encrypted with the same password.

    No script needed, just add rd.luks.name=<UUID1>=cryptroot1 rd.luks.name=<UUID2>=cryptroot2 rd.luks.name=<UUID3>=cryptroot3 rd.luks.name=<UUID4>=cryptswap to your kernel parameters and unlock all 4 with one password at boot.