HDD error

[norskman]

Is this terminal?

Do I need to change the HDD?

What is an offline error?

The following warning/error was logged by the smartd daemon:


Device: /dev/ada1, 24 Offline uncorrectable sectors

The status of the ZFS array is HEALTHY - so is this a drive replacement error or not?

I do have a spare.....the manual shows how to switch drives - I think the easiest way is to shutdown - remove the drive - restart so Freenas sees the failed array - All SATA using ZFS - then replace with the new drive and resilver.

Can someone confirm this is correct

 

[crisman]

The error messages mean the disk is failing, proceed with the instructions on the manual for disk replacement.

http://doc.freenas.org/index.php/Volumes#Replacing_a_Failed_Drive

 

[Ericloewe]

That drive is failing, replace it ASAP.

Follow the manual's instructions and you'll be fine. (I'm assuming you're running RAIDZ or mirrors...)

 

[SmallGuy]

First for the diagnostic, you can go further using the command

Code:

smartctl -a -q noserial /dev/ada1

And post the result in code tag.

 

[joeschmuck]

What you are going to find is drive ada1 has had 24 uncorrectable errors. This is one of the first indications your drive is failing, even though your pool is reporting healthy, which I'm sure it is for now. Follow the instructions Crisman linked to and by all means, shut down before pulling the drive. Many people try to hot swap a drive because it is technically supported by the MB, however it doesn't always work. Play it safe.

Now do you know which drive is ada1? type

Code:

smartctl -a /dev/ada1

and the serial number will be at the top. When you pull the drive, ensure you pull the correct serial number. Or you could use the GUI and click on Storage -> View Disks and record the ada1 serial number. Either is fine. Do not assume ada1 is the drive attached to the SATA0 connector.

 

[norskman]

ok, I now reloaded on to a new USB drive.
Then rebooted. All ok,

It sees my disks but not as the original ZFS - but as separate disks - ie my config looks like a new setup. Have I lost everything????

The good news the networking is ok again

 

[gpsguy]

try auto importing your volume

"If you click Storage → Volumes → Auto Import Volume, you can configure FreeNAS® to use an
existing software UFS or ZFS RAID volume. This action is typically performed when an existing
FreeNAS® system is re-installed (rather than upgraded)."

 

[norskman]

yes, I see that. How long does the import take?

It Identified my old ZFS setup. And is now auto-importing - its 4x3TB - so thats 9TB useable.

 

[gpsguy]

It shouldn't take too long. If it's doing the resilvering now, you can check it's progress by going to the shell and typing:

zpool status

 

[joeschmuck]

ok, I now reloaded on to a new USB drive.
Then rebooted. All ok,

What do you mean by this statement? You don't mean that your reloaded the boot drive do you?

 

[gpsguy]

Yes, the OP ran into a problem at boot time, after replacing the drive. See: http://forums.freenas.org/index.php...p-net-inet-tcp-delayed_ack.21606/#post-126236

I was confused too.

What do you mean by this statement? You don't mean that your reloaded the boot drive do you?

 

[joeschmuck]

Damn, that sucks!

 

[norskman]

ok, here is what I have done and the output.

I created a CD with Freenas Latest version 9.2.1.5 64 bit. Booted it up. Put a new USB 4Gb into the USB port - and loaded FreeNAS .That all worked.

I did use another USB stick - but kept the old one - I guess in case it was of any use?
I had not touched the HDD's from the previous set-up at all. In the home that the HDD and ZFS would still be there intact.
I now have the full GUI etc. View Disks shows the disks correctly. But no ZFS of course.
I tried to do an auto-import from the GUI.
That recognised my old ZFS volumes correctly - but never completes ( part of stage 2). Seems to hang - I waited overnight last night......This morning - nothing.
I tried to see what I could glean from the manual.
At the console CLI I typed: 'camcontrol devlist'
This showed me my four HDD drives, one Flash drive for the logs, and the USB drive. So looks ok.
'Zpool status' is accepted , but there is no output and then I cannot type anything more at the CLI. It never returns to the CLI. It just moves the screen up one line.

Of course I super worried about my data now. I thought if there is one thing that is safe its my FreeNAS. Never had an issue with it.

Are there any CLI commands I can use to get my ZFS reformed / recognised and get access back to my data? Even if I have to copy it off elsewhere?

 

[norskman]

I thought this through more - I replaced the failing drive - It was not dead - it is dying ( aren't we all...) and restarted. I then did an auto-import and it works! Phew.

Now I will go through the standard hard disk replacement.

I am going to read up on what system backups are available. It worries me that I could not restore my pool with a signle failed drive. Maybe the lesson is to have a second spare in the pool?
NAS storage these days is getting so large - few of us can afford to have 2 NAS systems replicating each other.

This is the ONLY problem I have ever had in 18 months running FreeNAS. It has been supremely fantastic and stable. What works does so very well. What does not is constantly being improved on.

I am a fan for sure.