ZFS HDD Mirror read error for one of two HDDs

Running a GPU and CPU stress test is a good PSU tester occasionally; I had a couple PSUs not handle that after a few years (fine for lower-powered stuff), and the first one I ever tried that on failed instantly :p (implying that might have been a random problem later)
 
HI as your HDD has no reallocated and no unrecoverable Sectors with only three Months Power on time , it appears to be healthy. A FEW DMA errors is not uncommon.
From a Population of 20000 HDDs we used to get half a dozen disks a Month with a 1000+ UDMA errors . That is when you know that there is a connection error between the HDD and the Motherboard.
 
Today after booting, and rebooting I got 2 errors.
1. Shutting down the system showed me the error: "Spin down disk failed" after issuing poweroff.
2. Booting up again showed me errors I previously posted, and if issuing zpool status I get the following:
Code: 
  pool: libraryz
 state: DEGRADED
status: One or more devices are faulted in response to persistent errors.
    Sufficient replicas exist for the pool to continue functioning in a
    degraded state.
action: Replace the faulted device, or use 'zpool clear' to mark the device
    repaired.
  scan: resilvered 1.59M in 00:00:01 with 0 errors on Thu Oct 23 14:31:20 2025
config:

    NAME          STATE     READ WRITE CKSUM
    libraryz      DEGRADED     0     0     0
      mirror-0    DEGRADED     0     0     0
        ada1.eli  ONLINE       0     0     0
        ada0.eli  FAULTED      0     0     0  too many errors

Trying smartctl -HA /dev/ada0 shows me this:
Code: 
smartctl 7.5 2025-04-30 r5714 [FreeBSD 14.3-RELEASE-p3 amd64] (local build)
Copyright (C) 2002-25, Bruce Allen, Christian Franke, www.smartmontools.org

Read Device Identity failed: Input/output error

If this is a USB connected device, look at the various --device=TYPE variants
A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options.

Could this still be a SATA cable error, or is it more like a HDD error now ?
If a disk fails to spin down upon shutdown, isn't it a HDD error ?
Kind of curious whether I should get new SATA cables now, or replace the HDD as I still have around 2 years warranty.

Another point I can think of is that the HDD was still doing some work, and the OS tried to forcefully close it, without asking the HDD to finish what it is doing.
I read that there are tools for Seagate HDDs to ask them politely to finish their task, before shutting down the system.
Just wondering whether there are similar tools for WD RED Plus HDDs.

Maybe I should try something like # camcontrol standby "device" before issuing poweroff ?
To confirm that the HDD is ready for a shutdown.

EDIT:
The HDD completely died on me.
It is not even recognized anymore by the BIOS.
Issuing zpool status gives me:
Code: 
pool: libraryz
 state: DEGRADED
status: One or more devices could not be opened.  Sufficient replicas exist for
    the pool to continue functioning in a degraded state.
action: Attach the missing device and online it using 'zpool online'.
   see: https://openzfs.github.io/openzfs-docs/msg/ZFS-8000-2Q
  scan: resilvered 1.59M in 00:00:01 with 0 errors on Thu Oct 23 14:31:20 2025
config:

    NAME                      STATE     READ WRITE CKSUM
    libraryz                  DEGRADED     0     0     0
      mirror-0                DEGRADED     0     0     0
        ada1.eli              ONLINE       0     0     0
        13283179534583615310  UNAVAIL      0     0     0  was /dev/ada0.eli

I am pretty sure it is the HDD now.
 
If it's still under warranty return it for replacement.
If it's out of warranty you can remove the hdd controller and check the contacts.
 
VladiBG said:
If it's still under warranty return it for replacement.
If it's out of warranty you can remove the hdd controller and check the contacts.
Click to expand...
Brought it back today, and in the following week I will get a new HDD.
Also handed over a smartctl log for the manufacturer to check and see whether he gets the same errors.
Pretty sure I will get a new HDD as replacement.

It wasn't the cable, because I even tried only one HDD at a time, and it failed with both SATA cables.
Once I get a new HDD I will see whether I can create a new pool as I messed up my old one, and all my precioius data needs to be gathered again, as my backup device is also broken and out of warranty :/
Bought a new backup device so, I will have fun getting my data back together...
 
hi, sorry to hear that. Nothing we could see from "smartctl" indicated that a total fault was about to happen.
As the media itself did not show excessive errors , I can speculate that onboard HDD controller was marginal somehow and failed .
ill be interested in the vendors response to this .
 
Now I am really confused.
My 1st HDD I sent to the repair service will come next week, but today after booting up the OS I noticed that the 2nd HDD I have now shows the same symptoms.
And the worst part is smartctl tells me that it cannot diagnose the HDD due to an Read Device Identity failed: Input/output error...
The device is listed as a /dev/ada0 in the device folder, though.
Yesterday it still worked fine.
I wonder if it is probably a SATA cable problem after all ?
Doing a dmesg -a | grep ada0 showed me the same errors as the 1st HDD had during boot.
 
It is always good to have extra sata cables on hand so you can test and eliminate the cable as the source of the problem.
 
depends on method , and internal sata port in the Dockstation , and booting capability,

If you have a DOCKstation into which SATA HDD can be installed , that has high speed USB ( eg 5 Gigabit USB speed ) and the motheboard
supports booting from USB , it may work. im not sure what bootstrap to use , as USB boot and SATA/HDD boot is not eqvavilent
 
I got the first corrupted HDD replaced.
The problem was that the NAS HDD is not designed to be in a PC which powers on and off frequently, and therefore the frequent boot up boot down process did wear the HDD faster than it should.
That was the explanation of WD.
Now I need to wait another month, before the second one gets replaced.

I was recommended to get a NAS server and just put the HDDs inside.
I don't know though which one I should get, and how could I set it up so it is always available 24/7 for my FreeBSD system ?
Like can I set the NAS server as a ZFS mirror RAID and mount that RAID to one directory ?
And if rebooting still access the content ?
I heard that I need to access the NAS through the network.
Quite new to NAS servers honestly so it would be my first one to set up.
 
You could just leave your FreeBSD server on all the time, and call it your "NAS server", which is a quite common thing to do.

i.e. FreeBSD + ZFS == NAS Server

Though, that 2000W power supply has got me a little concerned. The power bill may be an issue if you are using any significant part of that.
 
I wouldn't buy a Western Digital drive if they were the last hard drive maker around.

SDK Chan said:
I was recommended to get a NAS server and just put the HDDs inside.
Click to expand...
So this don't surprise me. Deflection is common on junk.

What I worry about is a refirb drive they will provide you with. I wouldn't have it in my system.

www.backblaze.com

Backblaze Drive Stats: Download Hard Drive Reliability Stats, Reports, and Test Data

Understand hard drive reliability and explore the hard drive test data from the Backblaze data centers. Backblaze is affordable, easy-to-use cloud storage.
www.backblaze.com www.backblaze.com
 
I'm the guy that had to tell customers all their lives photographs are gone when they bring me their WD Passport to revive.

If you shop for disks by lowest price you will regret it. With both disks showing trouble I hope you see this. The "NAS-Line" is just marketing term for talking heads.
They are not Enterprise Drives. Two new disks failing is not surprising to me from Western Digital.
 
SDK Chan said:
The problem was that the NAS HDD is not designed to be in a PC which powers on and off frequently, and therefore the frequent boot up boot down process did wear the HDD faster than it should.
That was the explanation of WD.
Now I need to wait another month, before the second one gets replaced.
Click to expand...
I have a problem with that explanation. It's bullcrap. No offense to you whatsoever. NAS drives should not die by startup count over 4 months.
These people would not admit if they were at fault if they had 1 billion drives returned.
I would be super cautious with replacement drives.
The factory has the ability to rest SMART when they certify it as a refirb. So you don't know replacement drives true mileage.
Refurbished drive in exchange for a four month old drive.
That's probably what you got. Sad to say.
Cut your losses with them. Go with an Enterprise Disk pairing.
 
Phishfry said:
I'm the guy that had to tell customers all their lives photographs are gone when they bring me their WD Passport to revive.

If you shop for disks by lowest price you will regret it. With both disks showing trouble I hope you see this. The "NAS-Line" is just marketing term for talking heads.
They are not Enterprise Drives. Two new disks failing is not surprising to me from Western Digital.
Click to expand...
That is the funny part, I paid a pretty amount for these drives.
Around $600 for both together.

Can you eventually recommend me something better then ?
Drives which will probably last for many years to come.
I looked at that Backblaze Drive Stat link you provided.
But according to your experience which company should I generally choose for HDDs ?

Phishfry said:
I have a problem with that explanation. It's bullcrap. No offense to you whatsoever. NAS drives should not die by startup count over 4 months.
These people would not admit if they were at fault if they had 1 billion drives returned.
I would be super cautious with replacement drives.
The factory has the ability to rest SMART when they certify it as a refirb. So you don't know replacement drives true mileage.
Refurbished drive in exchange for a four month old drive.
That's probably what you got. Sad to say.
Cut your losses with them. Go with an Enterprise Disk pairing.
Click to expand...
It is ok, because it is not mine explanation, but quoted from the WD staff which exchanged HDDs.
Thinking about it, I also came to the conclusion that it is bullcrap, but I didn't want to start a discussion right in the store.
I made good experience with WD back then, honestly, but that was now about 10 years ago.
My WD black nvme SSD also lives now around 7 years, and still works.
But in the HDD part they really dissapointed me honestly, and I even won't get the whole money back for that drives.
Knowing that they just sold me new faulted drives is kind of sad.
My first experience with such kind of treatment...

So, I basically should get "trusted and battle tested" HDDs and just put them in my PC case right ?
 
SDK Chan said:
and just put them in my PC case right ?
Click to expand...
Well how you install them can have an effect.

How tightly packed are they in drive hangers?
Leaving plenty of room around drives is preferable for airflow. Not stacked so tight the blow hole can't work.
Take a look at your SMART temperatures. Peak is important. If high add a drive fan.

I want to avoid recommending a brand. Even Toshiba now has "Enterprise" and "Enterprise Capacity" drive category.
Too many decisions...
 
we've been at this long enough to have heard terrible stories about every single drive manufacturer, so we don't buy into the "this manufacturer only produces bad drives" dogma that other people seem to be promoting.
 
The statistical averages will tell stories about some drive models and some manufacturers. So you can assess broad probabilities. Backblaze does that, very nicely. But statistics won't tell you where your drives will fall under the bell curve.

Buy "Enterprise Class" drives if you can afford them.

I would not power cycle a "NAS drive" frequently. The design assumption is probably that it will run continuously.

Do consider the operating environment of all drives carefully. Cooling matters, as does consistency.

I always configure my drives redundantly. I always have spares in store.

[I have 3TB WD Red drives (CMR WD30EFRX, not SMR) that have been running continuously since January 2013. Five have died in nearly 13 years -- one infant, and one young (both replaced under warranty), and two or three quite old (seven years, plus). Three are still running. However the WD Reds of today are probably very different to those of 2013, even if they have the same model number.]
 
Phishfry said:
Well how you install them can have an effect.

How tightly packed are they in drive hangers?
Leaving plenty of room around drives is preferable for airflow. Not stacked so tight the blow hole can't work.
Take a look at your SMART temperatures. Peak is important. If high add a drive fan.

I want to avoid recommending a brand. Even Toshiba now has "Enterprise" and "Enterprise Capacity" drive category.
Too many decisions...
Click to expand...
I would say there is a lot of room.
I have a very big tower with a lot of free room.
Temperatures were never really the issue as far as I monitored.
Well, I asked in other forums, and regarding WD HDDs I got similar answers, practically a "don't buy" recommendation.
Some people reported that Seagate lasted for them like 10 years or even more.
I never had Seagate HDDs before.


atax1a said:
we've been at this long enough to have heard terrible stories about every single drive manufacturer, so we don't buy into the "this manufacturer only produces bad drives" dogma that other people seem to be promoting.
Click to expand...
That is new to me.
Didn't know that such kind of territory existed.
There can be faulty drives, but getting 2 faulty drives in a row is questionable.


gpw928 said:
The statistical averages will tell stories about some drive models and some manufacturers. So you can assess broad probabilities. Backblaze does that, very nicely. But statistics won't tell you where your drives will fall under the bell curve.

Buy "Enterprise Class" drives if you can afford them.

I would not power cycle a "NAS drive" frequently. The design assumption is probably that it will run continuously.

Do consider the operating environment of all drives carefully. Cooling matters, as does consistency.

I always configure my drives redundantly. I always have spares in store.

[I have 3TB WD Red drives (CMR WD30EFRX, not SMR) that have been running continuously since January 2013. Five have died in nearly 13 years -- one infant, and one young (both replaced under warranty), and two or three quite old (seven years, plus). Three are still running. However the WD Reds of today are probably very different to those of 2013, even if they have the same model number.]
Click to expand...
My old 4TB WD Reds run like 11 years.
But the new ones, especially the Plus variant died in almost 6 moths / half of a year.

Once I get the second HDD replaced, I decided selling under warranty as they are.
Will probably try a Seagate Barracuda next (the green ones).
But if NAS really should run 24/7 then it is just a HDD not for me.
As I want to be able to power down the PC without the fear that after like 4 to 6 months an HDD breaks.
I can only say that probably the manufacturing of these WD RED HDDs was bad, or I had bad luck twice.
 
gpw928 said:
AFAIK, Seagate Barracuda Green drives spin down when not in active use. I'd seek advice before using them with ZFS.
Click to expand...
Do you mean they go into standby mode ?
Searched up, and Seagate Exos seems to be a good match, or IronWolf Pro if using them with ZFS.
 
SDK Chan said:
Do you mean they go into standby mode ?
Click to expand...
If you ask google: do seagate barracuda green disk drives use standby mode, it says "Yes, Seagate Barracuda Green drives use standby mode as a power-saving feature". So I would certainly seek advice before using them with ZFS (which generally has problems with drives that go into standby mode).
SDK Chan said:
Searched up, and Seagate Exos seems to be a good match, or IronWolf Pro if using them with ZFS.
Click to expand...
I don't have enough first hand knowledge of individual disk drives to have a good opinion. However Backblaze tells us that there's some 12TB drives worthy of consideration, e.g. HGST HUH721212ALE600 (12TB), Seagate ST12000NM001G (12TG), and Toshiba MG07ACA14TA (14TB). I see from previous stats that the WDC WUH721816ALE6L4 (16TB) is also very good. Note that all the Backbyte disk drives run continuously, and you need to check current availability.
 
You must log in or register to reply here.
Back
Top