Hello Forum,

 

Last week I came home to an unresponsive server. I believe my disk 3 is dead (showing "wrong" , I/O errors) but before I swap that out with the new one I want to make sure I'm doing the right thing. My parity is show up blue dot and I'm hoping to rebuild the troublesome drive from that.

 

Is the parity screwed? what should I do from here? I figure I should ask before I lose all my data. Syslog and screenshot attached. Version 5.0 beta14

 

Thanks to anyone / everyone for the help. 

syslog_050414.zip

So the blue dot means it's a new drive? It's not a new drive, been using it for over a year. Is there a way to mount it or trust it?

 

Is it just showing blue because one of the drives are broken/ gone/ unusable and needs to rebuild?

 

Any help would be great, thanks

Paste SMART reports for parity disk and disk 3.

Report for parity attached, just finished the extended test without errors (not sure how to export that to .txt doc). Disk 3 fails to create a report, I/O error (edit disk 3 attached, smart_dead.txt).

ParitySmart.txt

smart_dead.txt

I've come across this:

 

Yes, it's rather more complex then trust my parity. As far as I know, you can't put the array back into disk simulation mode because you will need a disk in the failed location. You basically need to replace the drive and then force unRAID to rebuild onto that drive.

 

1. Create an unRAID partition on the replacement disk.

2. Use the New-Config util to re-initialize the array.

3. Assign all the drives back to the proper disk spots and assign the replacement to the failed spot.

4. Type "mdcmd set invalidslot ##" at the command prompt, where ## is the disk number of the replaced disk. This tells unRAID which disk to rebuild.

5. Start the array.

6. I believe you also have to click a button to start the actual rebuild.

 

I'm not positive if #1 is required. However, there was a case or 2 on 4.x versions of unRAID where this was attempted without creating a partition and the rebuilt drive didn't have one after which required extra work to attempt recovery. You can do a new-config and asign just the replacement drive to say disk1 and start and stop the array and the disk will have the partition written.

 

It seems like this is the solution I need for my situation as well. Am I horribly mistaken?? Thanks!

I have a very similar situation right now and an open ticket on it: http://lime-technology.com/forum/index.php?topic=33250.0

 

I'm not rebuilding as yet, but will certainly follow this thread too.

 

EDIT: spelling

Bump - Will the process outlined above effect working disks 1 and 2?

This is a quote from Tom:

1. Execute Utilities/New Config.

 

2. on Main, assign all your drives, being very careful to assign Parity and your new disk (disk10) correctly.

 

3. From a console or a telnet session type this command:

 

mdcmd  set  invalidslot  10

 

(the 10 corresponds to disk10)

 

4. Click 'Start' on the webGui.

 

What should happen now is array gets started with disk10 reconstruct in process.

 

IMPORTANT: between steps 3 and 4 do NOT refresh your browser or navigate to any other pages in the webGui - just click the Start button which is already being displayed there (if you navigate to a different page or even refresh the browser after step 3 it will "cancel" the effect of that 'mdcmd' and result in your parity disk getting written - not good).

Just wanted to follow up with this. Thanks dgaschk, but unfortunately that didn't work for me. After following the instructions the system started to rebuild the parity off of the remaining 2 working drives.

 

A few notes that may help others.

 

My new drive was unformatted at the time, I have seen other threads the state the new drive needs an existing folder. I also didn't check trust parity. This seems like an obvious mistake now but at the time I wanted to follow the instructions posted as closely as I could.