Pool suspended after disk errors – how to safely recover data?

[sixdas]

Hello everyone,

I am running TrueNAS SCALE on a single-disk pool (unfortunately without redundancy). The disk is a WDC_WD40EFAX-68JH4N0 (4TB, ~5 years old).

Recently I received alerts about the pool being DEGRADED, and now it shows:

• Pool state: SUSPENDED – “One or more devices are faulted in response to I/O failures.”
• Disk is marked as DEGRADED.
• ATA error count has rapidly increased (currently 112).
• ZFS reports:
  • Read errors: 2
  • Write errors: 10
  • Checksum errors: 6

I already stopped all apps and services that were writing to the disk. A replacement disk (Seagate IronWolf 4TB) is arriving tomorrow.

My questions:

1. Is there a safe way to import the pool in read-only mode to copy as much data as possible before replacing the drive?
2. Can I still use the “Replace” function in the UI after inserting the new disk, or should I first attempt a manual data recovery?
3. Are there any recommended steps to avoid making the situation worse (e.g., scrub, resilver, etc.)?

At this point, my main goal is data recovery – I understand the risk of data loss with a single disk.

Thanks in advance for any guidance!

[joeschmuck]

It seems kind of funny, not in a good way, but I’m seeing a lot of people reporting single stripe vDevs having corrupt data. I would not say this is TrueNAS but rather many people who have been using TrueNAS without redundancy and the drives are starting to fail. Many people think the data is okay to fail and go away, until it happens to them, then they want the data back.

You didn’t have previous error messages?

ATA error messages can be caused by the SATA cable, power down, unseat and reseat the cable, power back online. to find out if the drive is actually failing, take a look at my link below to the Drive Troubleshooting Guide. It will help you know if your drive is failed/failing or something else.

ZFS errors does not mean physical drive failure, however it could be that, but never assume it is.

There are a few threads here dealing with similar situations, take a look for them.

[Sara]

Quick question, if it turned out to be a broken disk, would the simplest way not be to just create a mirror, resilver, remove the broken drive?

Or even better, get two good drives, make 3 way mirror and then remove the broken disk?