RAID 6 degraded

cjohnson19401 · 16. September 2016

This is on a 1.19 box. I've had this server up and running with an active clean RAID 6 array for quite some time now.
Just a few days ago I logged in to the webUI to apply some openssl/libc updates and I noticed that my RAID 6 volume was in a degraded state.....

Some basic disc info/mdad.conf etc. is here.....
http://pastebin.com/s37uqydz

Had to use pastebin because...

Code

Message is too long, must be below 10,000 characters.

Every 1TB disk you see there should be part of the array. The two other discs are a 320GB where OMV lives, and a 16GB flash drive that I use for flashing RAID firmwares, boot recovery, etc..

Obviously, I'd like to get the array back to a clean state. Why isn't it showing me which disk failed? How can I go about replacing/investigating the failed/missing disk?

ryecoaaron · 16. September 2016

The output of cat /proc/mdstat does show you the "good" disks. Just compare that with the output of fdisk and see which one is missing. If it doesn't show up in fdisk, the drive has failed. With that many disks, you will need to go through the physical devices tab and compare serial numbers.

cjohnson19401 · 16. September 2016

I'm used to seeing an "(F)" next to the sd name output in mdstat. Hmmmmm.....

ryecoaaron · 16. September 2016

I don't see the failed flag on any of them in your pastebin. Did it just show up? That would definitely make it easier.

cjohnson19401 · 16. September 2016

No. It didn't show up. That's exactly what's weird about this. I am used to seeing some sort of flag next to the sd name to indicate which disk is missing/removed/failed etc....

Not sure why it's not there now?

ryecoaaron · 16. September 2016

If the drive completely fails, it won't show up with the (F) flag because the system doesn't know what device it is.

Jetzt mitmachen!