Hi all,

here is the problem: My config was two data drives and one parity (and two cache drives). One of the data drives was failing and I wanted to check on it. During this process my flash drive also died. To make things worse, I totally forgot to backup my flash drive. (That's something that is never going to happen to me again!)

So now, created a new flash drive, updated the key and identified my drives. Is there a way to set up the array in a way that saves my data?

I was thinking about 1. assigning the drives and starting the array with 'parity is already valid' and 'maintenance mode on', 2. turning array off and 3. starting array regularly. Does that make sense? Thank you all in advance!

 

When the flash drive failed was the failing disk disabled or still enabled?

I didn't disable it. I hope, that is what you ask?

Edited March 9, 20242 yr by MrSatchmoo

Not really, I just wanted to know if the disk was disabled or not, it could have been disable by Unraid, did all disks have a green dot?

I think it showed something like not connected

It would be good to know for certain, but if you don't, post the diagnostics first, after assigning the drives as they were, but before starting the array.

Thank you so much for your help!

What part of the diagnostics file should I post? The whole zip file seems a little bit to much.

Btw, I actually have a diagnostics zip file from between hdd and the flash drive failure. Can I find anything helpful there?

1 hour ago, MrSatchmoo said:

Thank you so much for your help!

What part of the diagnostics file should I post? The whole zip file seems a little bit to much.

Btw, I actually have a diagnostics zip file from between hdd and the flash drive failure. Can I find anything helpful there?

We normally want the whole zip as that stops us making continual requests for more of its contents.

sure thing! here are both.

tower-diagnostics-20240213-1757.zip tower-diagnostics-20240310-0859.zip

The earlier diagnostics shows that your drives were:

  Parity: WDC_WD40EFRX-68N32N0_WD-WCC7K2TTPE1K

  Disk1: WDC_WD40EFRX-68N32N0_WD-WCC7K6RETKX4

 

so I assume that the other drive (which was already disabled even then) was

  Disk2: WDC_WD40EFPX-68C6CN0_WD-WX22D83RHEP8

although it was showing as not online at the time, but was being  emulated so any new writes were only being reflected in parity - not in the physical disk.

 

Have you done a New Config yet as the last diagnostics look like might you have?  If so the next steps are likely to be:

• Tick the Parity is Valid checkbox
• Start the array in Maintenance mode
• Stop the array
• unassign disk2
• start the array in normal mode to get disk2 emulated
• post new diagnostics

It is probably also worth seeing if the physical disk2 can be mounted in Unassigned Devices as that would tell us if there is a fallback position if the emulated disk has problems

 

 

 

 

 

Hi,

the old (failed) drive (a Seagate) does not show anymore. or rather: As soon as I try to assign it to the array, it vanishes. Because of that I removed the Seagate from the server altogether. The WX22D83RHEP8 is the new one I bought to replace the Seagate.

Should I try to reinstall the failed drive again, before starting the Steps you mentioned?

Thanks again!

No.   The point of the steps I mentioned was to see if you can successfully emulate drive2 as all the rebuild process does it make  a physical drive match an emulated one.

Thank you! So, I'll try to summarize what I understood so far about my situation, just to be sure, that I'm on the right track here.

If I'm lucky, I have a valid parity disk and one of two data disks. Now I need to tell the system to recreate the second data disk from the before mentioned two disks. As my configuration from my old setup is gone due to the failed flash drive, I somehow have to make sure, that unraid knows wich one of the data drives is valid and wich one is new/replaces the broken one, so it rebuilds that one and not the one with my data on it. Am I correct so far?

Question is, how do I do that. 😨

11 minutes ago, MrSatchmoo said:

Thank you! So, I'll try to summarize what I understood so far about my situation, just to be sure, that I'm on the right track here.

If I'm lucky, I have a valid parity disk and one of two data disks. Now I need to tell the system to recreate the second data disk from the before mentioned two disks. As my configuration from my old setup is gone due to the failed flash drive, I somehow have to make sure, that unraid knows wich one of the data drives is valid and wich one is new/replaces the broken one, so it rebuilds that one and not the one with my data on it. Am I correct so far?

Question is, how do I do that. 😨

You follow the steps I gave earlier.   Make sure that the parity drive and disk1 correspond to the serial numbers I gave for them.

here we go.

tower-diagnostics-20240310-1420.zip

Check filesystem on the emulated disk2, run it without -n, and if it asks for -L use it.

Sorry for stupid questions, but how do I do that? according to the manual I need to start the array in maintenance mode. is that correct?

Right now disk 2 shows as not installed / unmountable and I see nothing to do a check.

Stop the array, click on the emulated disk2 and set the filesystem to xfs, start the array in maintenance mode and check filesystem

OK, done. should I upload a log file?

edit: I did the check without -n. It didn't ask for -L.

Here's the data that showed on screen.

Thank you both again!! Awesome!

check_filesystem_status.txt

Edited March 10, 20242 yr by MrSatchmoo

Start the array in normal mode now, the disk should mount, look for a lost+found folder.

Again, thank you, and, again, sorry for more questions. But just to be sure, normal mode with parity and both data disks in the array?

Don't change anything in the array, just start in normal mode instead of maintenance mode, and post new diags.

here is the file and a screenshot of the lost & found folder.

tower-diagnostics-20240311-1724.zip

Sorting out the lost+found folder is unfortunately a manual task.  You have to decide if it is worth the effort.  The Linux 'file' command can at least help by giving the content type of each file.

If you still have old disk2 intact, and it's not completely dead, it may be worth see if you can copy some data from there.