Here's the situation.  Running Version 4.4.2.

 

I recently had a drive fail and was just planning on replacing it.

 

First off, I accidentally hit Restore and then start.  I stopped the array after I discovered the Advanced Format drive postings after only about 5 minutes.  So I then removed the drive and then added the jumpers and replaced the drive.  I then unknowingly again hit Restore and then start.  I noticed the Parity drive was writing which I didn't think should be happening. So I stopped it again after about 5 minutes.

 

I just realized also that my Parity drive is also an Advanced Format drive without any jumpers.

 

So my question is, how do I get my unRaid back into working condition and prevent any more data loss and restore the failed drive as well as not lose anything on the other 4 drives in the system?

 

I'm assuming once I get the unRaid working again, I should probably upgrade the version.  Thanks for any assistance.

syslog.txt

Here's the situation.  Running Version 4.4.2.

 

I recently had a drive fail and was just planning on replacing it.

 

First off, I accidentally hit Restore and then start.  I stopped the array after I discovered the Advanced Format drive postings after only about 5 minutes.  So I then removed the drive and then added the jumpers and replaced the drive.  I then unknowingly again hit Restore and then start.  I noticed the Parity drive was writing which I didn't think should be happening.

 

I just realized also that my Parity drive is also an Advanced Format drive without any jumpers.

 

So my question is, how do I get my unRaid back into working condition and prevent any more data loss and restore the failed drive as well as not lose anything on the other 4 drives in the system?

 

I'm assuming once I get the unRaid working again, I should probably upgrade the version.  Thanks for any assistance.

You've basically done everything possible to eliminate any possibility of getting to the data on your failed drive.

 

Consider it gone.  Think strongly of the words "other backup copy"

 

Oh yes, the "Restore" button effectively sets a new fisk configuration, and immediately invalidates parity.  You have no way to recover your old data unless the failed disk magically comes back to life.  Adding jumpers to a drive with data is the other kiss-of-death.  It made those disks un-readable.

 

to sum it up

 

The data on the parity drive is invalid.

The data on the failed drive is gone.

The data on any drive where you removed or added a jumper is in-accessible and probably gone too. (You MIGHT be able to remove the jumpers and recover the data... probably worth a try)

 

Sorry.

 

You can only now stop the array.

Assign drives as wanted

Press "Restore" to set a new disk configuration based on the remaining working disks

Start the array to have it completely rebuild parity as if you were doing it for the very first time.

Re-format the disks that show as un-formatted.

 

Read here about the Evils of the Restore Button http://lime-technology.com/wiki/index.php?title=Best_of_the_Forums

 

Joe L

 

The restore button did not restore data, it restored your configuration to a brand new state, immediately losing parity protection and the ability to rebuild a disk.  There are ways to recover from hitting the restore button, but If you hit start after you pressed restore, the parity data is rewritten to match the configured drives and your ability to recover completely lost.

 

Your best option would be to access the dead or dying drive to pull the data off, if you can get it to work for a while. Search google for info about putting hard drives in the freezer to bring them back to life long enough to do a recovery.

 

If you had started the array but quickly realized and stopped the parity build, there are ways we might have tried to recover some of your data, but if you let the parity build for a long time, your results would be poor.

Use the trust my array procedure in the wiki except substitute the disk number for the 99 in the procedure. You should see the replacement disk having writes when you start the array. Don't get the disk number wrong or you'll damage data on another disk.

 

http://www.lime-technology.com/wiki/index.php?title=Make_unRAID_Trust_the_Parity_Drive,_Avoid_Rebuilding_Parity_Unnecessarily

 

Then, run the reiserfsck checks as outlined in the wiki.

 

http://www.lime-technology.com/wiki/index.php?title=Check_Disk_Filesystems

 

Say the parity build takes say 6 hours or 360 minutes. You just lost 5 minutes / 3600min * 100% = 0.14% of the data that is on the disk.

 

I believe that during the disk rebuild the disk will appear as unformatted and that you have to hit the format button.

 

Others have recovered a lot of data after doing a similar thing.

 

Peter

Let me just clairfy since the responses are referring to the "failed" drive.  The drive I'm referring to in my first post is an empty replacement drive.  Does that change anything?

 

If I go to my devices tab...the first four drives are still all listed.

 

You are saying that even though the data is sitting on the Parity drive, it is unrecoverable...correct?  I'm assuming I can't pull the parity drive out and at least recover the data in a Windows 7 machine...correct?

Your parity drive was rewritten when you started your array after pushing restore. You said you stopped it after 5 mins but then pushed restore and started the array (and the destruction of your old parity data) again. You never mention stopping it again, so I am assuming it is still running or it ran to completion. If it is still running, you should stop it!

Let me just clairfy since the responses are referring to the "failed" drive.  The drive I'm referring to in my first post is an empty replacement drive.  Does that change anything?

 

If I go to my devices tab...the first four drives are still all listed.

 

You are saying that even though the data is sitting on the Parity drive, it is unrecoverable...correct?  I'm assuming I can't pull the parity drive out and at least recover the data in a Windows 7 machine...correct?

There is no "data" on a parity drive.

You said you were replacing a failed drive.  That data is now effectively gone.   You specifically invalidated parity... twice.  You overwrote about 5 minutes worth at roughly 60MB/s.  That is highly likely to cover the initial part of the main file system structures on the failed disk.  

 

Right now the syslog shows a missing disk5.

 

Best you can do right now is put in a replacement for disk5 and let it rebuild.  It might be usable data, probably not.

 

Keep your finger off the restore button.  It was removed because of situations like this.

 

Joe L.

Your parity drive was rewritten when you started your array after pushing restore. You said you stopped it after 5 mins but then pushed restore and started the array (and the destruction of your old parity data) again. You never mention stopping it again, so I am assuming it is still running or it ran to completion. If it is still running, you should stop it!

 

I actually stopped it both times after about 5 minutes.  It didn't run very long.

Just to provide more details.

 

I just want to make sure I don't loose any data on drives 1-4.

My 5th drive is the one that failed.

1) I removed the failed drive, inserted the replacement empty drive.

2) Hit Restore, then start, stopped after about 5 minutes

3) Unassigned Drive 5

4) Removed the 5th empty drive and added jumper to pin 7&8 on the replacement empty drive

5) Inserted the replacement empty Drive again and assigned back to slot 5

6) Hit Restore, then start, stopped after about 5 minutes

 

Peter - do you foresee any risk in me loosing data on the other four drives by doing the procedure you recommended?

Joe - are you saying that you don't think Peter's method will work?

 

I still have the failed drive sitting here.  Could I assign that to slot 5 even though it's giving errors?  If so, could I just hit START after putting it back in?

if you were to assign the old failing disk to slot5 now it would be treated as a new drive and an attempt would be made to clear it. You erased the arrays knowledge of it when you pressed the botton labeled as restore.

do not assign it to the array unless ypu want it cleared.  Not withot very special commands anyway..

 

 

So I was able to put the failed drive back in just so I could see what was on it and I don't think it's a great loss.

 

So now I'd just like to get my unRaid working with the four remaining data drives so I can upgrade to the later version of unRaid that supports the newer Advanced Format drive.

 

I unassigned the failed drive and will remove it.

 

But now do I hit RESTORE and START and let parity-sync go through the process to make sure my system is protected?

If your old disk 5 is working fairly well for short use, I think I can give you steps to fully recover it. If you don't care about it's contents, you can create a new array with your dour disks. If you want me to document the steps, let me know and I will do this evening.

If your old disk 5 is working fairly well for short use, I think I can give you steps to fully recover it. If you don't care about it's contents, you can create a new array with your dour disks. If you want me to document the steps, let me know and I will do this evening.

 

I was only able to view folders on the drive...nothing seemed to be in the folders.  So I'm not sure how recoverable it will be.  Your thoughts?

 

If I do need to resort to going down to the four drives, do I hit RESTORE or just hit START?

Peter - do you foresee any risk in me loosing data on the other four drives by doing the procedure you recommended?

 

As long as you use 5 in the command for disk5 it won't hurt the other disks.

 

Otherwise, unassign disk5 (failed or new, doesn't matter) hit restore again and start. Parity will rebuild and disk5 will be completely forgotten about.

 

Peter

 

Take your time. Evaluate the options. Then do what you think is best.

Ok - Guys. Back again in need of assistance.

 

I got the four drives working back again and then upgraded to 4.7 version by copying the necessary files to the Flash drive.

 

I placed in drive 5 again and it doesn't show up in the list as an option in the Devices tab drop downs.  Any ideas?

 

I've also tried putting a new Seagate 2TB AF drive and it didn't see that one.

Forgot to attach the Syslog.

 

syslog.txt

Any insights into the drives not showing up since upgrading to 4.7?

 

My unRaid has been down for over a week so I'm anxious to get it back working again.

Anyone know how to contact LimeTech via phone?  I'd love to figure this one out. 

 

Are there any issues with rolling back to 4.4.2 so I can add these drives to the unRaid?

I just restored my version back to 4.4.2 and the drive becomes assignable again.

 

So I'm at a crossroads as to whether to just stick with this version or figure out why the drives aren't assignable when upgrading to 4.7.  I've scoured the forum in hopes of finding an answer but no luck.

 

Has anyone experienced this same issue and if so, what did they do to get 4.7 to see the drives?

 

I'm definitely surprised to hear that the drives don't appear in 4.7 but do appear in 4.4.2.  I figured it would be the other way around.

 

Are the new drives connected to a SATA controller or directly to the motherboard?  Have you precleared the new drives to ensure that they are trustworthy?  Do you trust all the SATA cables/connections?

I'd have to pull it out of the rack in order to find out for sure, but it's most likely connected to a controller.

 

I'm pretty confident that it's not a cable issue since I left the drive in the same bay when going from one version to the next during my testing just to verify.

 

BTW, I tried to upgrade to 4.6 and had the same issue of the drives not being assignable...so that may narrow the search from something that changed from 4.4.2 to 4.6.

 

I haven't precleared any of the newer drives yet.  Do I have to do that first before they would show up as assignable?

 

As a side note, the drive I inserted was one that was recently removed due to errors.  So I'm not sure if that would give a false positive since it was in the unRaid before.  I was reluctant to put in one of the newer advanced format drives until I figure things out further.

Ok.  I want to try and prevent any issues before I create them.

 

I am still running version 4.4.2

 

I have another "disk7" that is a Western Digital 20EARS that is showing a red dot.

 

I have a 2TB Seagate Barracuda LP that I was going to replace it with.  Do I need to worry about a jumper on the Seagate replacement?

 

Is this the correct method to replace it:

1) Stop Array

2) Power Down

3) Swap Bad Disk with Seagate

4) Configure the new Seagate as Disk7

5) Start Array

 

Only EARS used a jumper with pre-4.7 versions. Seagate has some magic called Auto-align.

Is it normal that I've been trying to replace a failed disk and all disks have the same number of errors and fail at data rebuild around 65%.

Both 2 TB drives fail with 288 errors. This is the third drive I've tried to replace the failed one.

 

I'm wondering if it's possible that it's not the drives and something else.