Tuesday 22 Jul 2025

Increasing Redundancy in My ZFS Backup Pool

I decided to increase the redundancy of my ZFS backup pool, which stores my entire digital life—photos, videos, music, and media. My current setup consists of three 4TB hard drives in a RAIDZ1 configuration (similar to RAID5). This means that I can afford to lose one drive without losing data. However, if the rebuild fails and another drive dies, I would lose everything on this pool. I have backups elsewhere but would prefer to not spend hours retoring.

To mitigate this risk, I wanted to upgrade to a two-drive redundancy setup using a spare 4TB drive I had. However, this meant I had to destroy my existing pool and start over from scratch. Here's how I did it without losing data:

Step 1: Prepare the New Drive

First, I added the fourth 4TB hard drive to my PC. Next, I removed one of the existing drives from my pool. At this point, the pool wasn’t degraded (which is okay because I had backups).

Step 2: Create a New RAIDZ2 Pool

I tried to create a new RAIDZ2 pool using two of the drives. However, RAIDZ2 requires a minimum of four drives, so I had to create "fake" disks during the setup process. Here's the command I used to simulate the necessary disks:

truncate -s 1G /mnt/disk1.img
truncate -s 1G /mnt/disk2.img

Once the pool was created, I exported it and removed the two fake disks. After that, I imported the pool back into ZFS, which now showed as "degraded."

Step 3: Copy Data to the New Pool

To transfer my data, I used the zfs send command to copy the datasets from the old pool to the new one. This took a long time, but eventually, everything was successfully transferred.

Step 4: Destroy the Old Pool and Attach Drives to the New Pool

Once the data was moved, I destroyed the old pool. I then attached one of the old hard drives at a time to the new RAIDZ2 pool, waiting for the rebuild to complete before attaching the next drive.

  pool: rpool2
 state: DEGRADED
status: One or more devices is currently being resilvered.  The pool will
	continue to function, possibly in a degraded state.
action: Wait for the resilver to complete.
  scan: resilver in progress since Sat Aug 16 10:34:30 2025
	313G / 5.94T scanned at 5.69G/s, 0B / 5.94T issued
	0B resilvered, 0.00% done, no estimated completion time
config:

	NAME                                        STATE     READ WRITE CKSUM
	rpool2                                      DEGRADED     0     0     0
	  raidz2-0                                  DEGRADED     0     0     0
	    ata-Drive1				    ONLINE       0     0     0
	    ata-Drive2				    ONLINE       0     0     0
	    ata-Drive3				    ONLINE       0     0     0
	    replacing-3                             DEGRADED     0     0     0
	      1826073647085818739                   UNAVAIL      0     0     0  was /mnt/disk2.img
	      ata-Drive4			    ONLINE       0     0     0

Final Setup: A Secure RAIDZ2 Pool

Now, I have a RAIDZ2 pool that allows for the failure of two drives, which significantly increases the redundancy and security of my backup server. This new configuration also gives me more confidence when using older drives, as the system is now protected against multiple drive failures.

  pool: rpool2
 state: ONLINE
  scan: resilvered 1.45T in 03:49:32 with 0 errors on Sat Aug 16 14:24:02 2025
config:

	NAME                                      STATE     READ WRITE CKSUM
	rpool2                                    ONLINE       0     0     0
	  raidz2-0                                ONLINE       0     0     0
	    ata-Drive1				  ONLINE       0     0     0
	    ata-Drive2				  ONLINE       0     0     0
	    ata-Drive3				  ONLINE       0     0     0
	    ata-Drive4				  ONLINE       0     0     0