Help me understand redundancy

I finished creating a RAIDZ2 pool and was curious to know about the redundancy and how does it actually work.

If i’m not wrong, RAIDZ2 requires 4 disk and provides upto 2 disk failures. Let’s say i’m using 8x10TB disks. Out of which i’ll get 6x10TB as 2 will be used for partity. Roughly, 6x10TB = 60TB of raw capacity and let’s say the pool has 40TB data. So, how come just 2 parity disks which just has 20TB raw capacity protects the data when it’s actually 40TB of data?

You are using two disks of space for parity, not two specific disks. It’s spread over all the disks in the VDEV

White Papers | TrueNAS - Open Enterprise Storage Look for Pool layout whitepaper at bottom

1 Like

…and while ZFS’ implementation differs considerably from earlier versions of RAID; the principles of parity RAID go back decades. If you want to understand the details of how it works, old documentation on RAID 5 and RAID 6 should also be relevant.

2 Likes