Volume degraded - but drive seems ok

[Junicast]

Hi,

I got surprised about a degraded volume.

This is the output of zpool status:
http://paste.debian.net/714559/

Those are my devices:
http://paste.debian.net/714560/

I can see all 5 disks there were initially.

Please also see attached screenshots.
I'm running FreeNAS-9.10-STABLE-201605240427 (64fcd8e).

To me it seems like that my disk received some kind of new ID.
I'm unsure about what to do next. Please help!

 

[Bidule0hm]

Can you post the list of your hardware please?

 

[Junicast]

Have you spooted my screens? What else do you need?

 

[Bidule0hm]

Yes and they don't list the MB model, if you use an HBA and what its model is, the PSU model, ...

 

[Junicast]

Here's my dmidecode output:
http://paste.debian.net/714600/

It's a Gigabyte P67-ud3-b3 Mainboard, no HW Raid controller, all drives onboard connected.
PSU is Pure Power CM L8 530W bequiet.

 

[Junicast]

Well that's confusing. In Volume status I am able to replace the missing drive with the one that's been used for that Raid-Z before.
But when I try to replace it gives me a warning about a found ZFS partition. This is of course true, because it's the actual drive

Is my assumption right that my best shot would be to rebuild / resilver with that very drive and just hit the force radio button (see screenshot).

 

[Bidule0hm]

Ok, so here are my comments:

- the PSU isn't the best in the world but is pretty good, good enough to not cause a drive drop at least (unless it's several years old I guess).
- you use non-ECC RAM. You may want to check it with Memtest86 to see if there's any problem.
- the MB is desktop consumer grade, not server grade. The biggest bad thing on it is the Marvell SATA controller, there's a ton of problems all over the forum related to Marvell controllers IIRC.

So my conclusion is that the Marvell controller is causing the drive to drop from the array.

Any SMART info on the drive? (use smartctl -a /dev/adaX with X replaced by the correct number for the drive, and post the output on pastebin or in codes tags so it's readable)

 

[Robert Trevellyan]

Looks like your pool is encrypted. Make sure you follow the documented drive replacement procedure very carefully.

 

[Junicast]

Thank you all.
My PSU is fine, I checked it, also ran memcheck for a couple of days just recently.
Thank you about the controller hint. I will check if there's something better. So far I haven't had any problem for more than a year. Performance is also ok.

In the end I actually did the replace procedure given in the GUI. All went fine. RAID Volume is ok again now and I also did the re keying and also backed up the keys.