Help with replacing erroring disk

TrueNAS General
1

Hello,
I am truenas and zfs newbie running a small nas server on a shoestring budget. It has 3x1TB hard drives configured in a raid RAIDZ1 configuration. 1 of my hard-drives started showing errors on the weekly SMART tasks and throwing error alerts since past 2 weeks. The latest one is:
Device: /dev/sdc [SAT], 3288 Offline uncorrectable sectors (changed +8).
Device: /dev/sdc [SAT], ATA error count increased from 96 to 106.

I guess i should replace that disk ASAP. I have 2 questions regarding replacing the disk.

  1. Should/can I only replace the 1 bad disk new 1TB hard disk only (the same size), can I replace it with a larger capacity disk like 4/6 TB. That way i can slowly upgrade the pool capacity in the future. If it is possible to go with a upgrade with larger size disk how will RAID configuration and size of the pool be.

  2. Are there any instructions which lists the step to replace a bad disk from ZFS pool. Like should i do some prep steps ? should I shutdown the NAS to replace disk (i have hotswap bays) or do it live ? how long does it take to recover etc. ?

Any other advice would be greatly appreciated.

Thank you

System Information
ElectricEel-24.10.2
Model:Intel(R) Xeon(R) CPU E5-2650 0 @ 2.00GHz
Memory:63 GiB

2

My ZFS pool status is healthy
Pool Status:Online
Total ZFS Errors:0
Scheduled Scrub Task:Set
Auto TRIM: Off
Last Scan:Finished Scrub on 2025-03-02 00:06:31
Last Scan Errors: 0
Last Scan Duration: 6 minutes 27 seconds

3

Okay I got the link to disk replacement page, soon after I posted the query, thanks

But still wanted advice on question 1
Thanks

4

Yes, you can replace with a larger disk. And yes, you can slowly replace the others as they die or you can afford them. Even totally mis-matched. Like 4TB for this time, and 6TB next time.

The way this is handled, is that you have to replace all the disks in a vDev before that vDev with grow to allow use of the new space that is common to all disks. A ZFS vDev is like your 3 disk RAID-Z1, which acts like a virtual disk device, (aka vDev). Other vDev types like Mirrors or dRAID exist. ZFS’ redundancy is not at the “pool” level, but at the vDev level, like your 3 disk RAID-Z1.

One note. If you have a spare disk slot for the replacement disk, you can do a “replace in place”. Meaning, if you need to replace a failing, but not yet failed, disk, it can be used to maintain redundancy during the re-silver to the new, replacement disk. Kinda like Mirroring the failing disk, and when done, ZFS breaks off the failing disk.

This becomes important for RAID-Z1 and 2-way Mirror vDevs that only have 1 disk of redundancy. Removing the failing disk for replacement means your vDev / pool has no redundancy.

3 Likes