After starting my Unraid 3 Tower which has 2 4tb Parity Disks which had passed a parity check about a month ago and I see it is showing 2 disks Emulated with Disk 2 not showing at all & Disk 5 with CRC errors and doing its own self check/rebuild. Both are 2tb and realatively full at highwater 250gb free on both.

I stopped Disk 5 & because it was easier to get to I replaced with a 4tb & reattched disk 2 (which I believe was a mistake) only because Disk 2 was the the only one to be rebuilt (I still have the faulty Disk 5). Therefore it is best to rebuild only 1 disk at a time even though Disk 5 was getting written to it.

I am considering putting the faulty disk 5 back in then see if I can get it to rebuild CRC itself again so that later on I can recover that data in a New Config. Is this the best way to go?

The last time something similar happen like this, it was a Parity disk 4tb & a 3tb data disk but I was able to recover both onto 2 4tb drives I had spare, I am sure it fixed the Parity 4tb 1st & then I did the data disk later. But this time I panicked a little.

Your diagnostics might show what happened if you didn't reboot. In any case, it would provide context to understand the situation better.

Go to Tools / Diagnostics and attach the full zip to your next post.

8 hours ago, redlaws said:

After starting my Unraid 3 Tower which has 2 4tb Parity Disks which had passed a parity check about a month ago and I see it is showing 2 disks Emulated with Disk 2 not showing at all & Disk 5 with CRC errors and doing its own self check/rebuild.

It won't initiate a rebuild on its own. And it can't do a parity check with 2 disabled disks, but it can do a read-check. I don't know if unclean shutdown will start a read-check in that situation or not.

 

Go to Tools - Diagnostics and attach the complete Diagnostics ZIP file to your NEXT post in this thread.
 

2 hours ago, trurl said:

I don't know if unclean shutdown will start a read-check in that situation or not.

I think it will   There is no explicit command to do a read-check as far as I know.   You just ask for a check to be started and if you have no current redundancy then a read check is carried out.

I probably should let the original disk 5 go through with what it was rebuilding (but I could only see on the server monitor and didn't know how long that would take or if it would fix itself) but with disk 2 out I was concerned.

Here is what I collected at the time but what I would like to know is "can it rebuild 2 data drives at the same time".

Also the 4tb (dom 2020) that I put in as a replacement for disk 5 probably had  a similar error when that failed (Parity a month or so ago) & it needed to be reset with the connections on the drive but I felt replacing both drives & rebuilt parity then the other drive afterwards was a safer bet.

 

This is my biggest server out of the 3 I have, it might be time to strip it down & clean it up & I have a spare 8way card I can put into it but it did have about 10x 2tb I wanted to replace at some point to 3tb or 4tb drives

cylon-diagnostics-20210816-2028.zip stower03-64-smart-20210822-1157.zip

This was a SATA controller problem, quite common with some Ryzen boards, multiple disks dropped offline, reboot so they can come back online, start array and post new diags.

Your syslog is so full of controller issues can't really tell what is going on. Not clear that anything is being rebuilt. You didn't specifically mention that you had done anything to make it begin rebuilding, in fact, your wording seemed like you though it had decided to do a rebuild itself, which it won't do. So maybe it is only doing a read-check.

 

In any case, no point in continuing until you fix your hardware.

I have finally found some time to look at this server again, I have put the old 2tb Disk 5 (removed the replaced 4tb) back to see if I can recover what is on it. The only way I can see to do that is a new config (assuming that everthing is readable), apart a something that can read xfs & copy the data off it (1.63tb).

I think putting the 4tb back & see if I can check the xfs system & then able to read what is on the 2tb with Ubuntu virtual machine so that I may be able to copy off any data on it.

 

So with 2 parity drives ok & 2 failed that are both Emulated what is the best or prefered way to fix them reattached both & rebuild them or shutdown & replace both drives one at a time or both together.

I believe the controller card errors may date back when I purchase 3x 8 way controller cards from ebay, because when I was testing them I place 1 card into this server at a time with the sas leads attached to see if they worked in unraid bought 3 different card that appear to have a jbod that matched best matched the Super micro cards I already had.

I will look at them & put together what I found.

when I starting up the Server & could not get the windows page up but the monitor connected showed something like this.

 

 

similar error.pdf

12 hours ago, redlaws said:

The only way I can see to do that is a new config (assuming that everthing is readable), apart a something that can read xfs & copy the data off it (1.63tb).

I think putting the 4tb back & see if I can check the xfs system & then able to read what is on the 2tb with Ubuntu virtual machine so that I may be able to copy off any data on it.

You should be able to mount and/or repair the disk as an unassigned device. Can't see any reason to get a VM involved.

thanks trurl for that suggestion I will try that, I am not sure if I have many/any spare sata ports but it may be time to clean it up a bit & put in another card. Also clean up the usb drive too.

On 10/10/2021 at 5:16 AM, trurl said:

You should be able to mount and/or repair the disk as an unassigned device.

I have been pondering this comment for a while now as I am hopeful that I still have data on my old 2tb disk 5.

one option was to copy the setup of the usb to another usb & add a 4tb parity & just try to a new config with those 2 drives & then once parity is correct for that replace my 4tb disk 5 in as a rebuild, so that then I can re-add it back to my server & rebuild parity again.

is this a quicker way to fix my missing data on disk 5 (4tb).

 

If the disk is unmountable, no reason to expect rebuild to fix it. If you just want to put it in a system by itself to try to repair its filesystem, no need for parity. If you have a spare port on your Unraid, no need to get another system involved, just use Unassigned Devices to check filesystem on it.

thank you for that, I am hoping that I can find the easiest & quickest way to hopefully put the 1.63gb of data on the 4tb hdd from the old 2tb.

I have only just updated this array to 6.11.5 from 6.9.2.

I had disks 5-7 extra 4tb play up so I built a new config & parity & then rebuilt all 3 with new disks.

I then put the original disk 5-7 back into tower 3 and was able run a parity check this year.

I had started to delete and clean up some data and wanted to update parity disks, for disk 5 & 7 to to give me a Red X.

I have played around with the connections for disk 2, 5 & 7.

I can afford to lose disk 5, 6 & 7 from this array. Should I remove 5 & 7, but can I rebuild disk 2 or the parity disks not complete.

I might take disk 2 & add it to my new array & rebuild onto a new disk.

 

Please post the diagnostics.

9 hours ago, redlaws said:

can I rebuild disk 2

On 1/2/2023 at 11:31 AM, trurl said:

If the disk is unmountable, no reason to expect rebuild to fix it.

 

3 hours ago, JorgeB said:

Please post the diagnostics.

 

here they are

stower03-64-diagnostics-20230503-2314.zip

Check filesystem on disk2.