So the short story is that my parity drive kicked itself out of the array (write to it failed I know for sure).  I ran a short smart test on it and it found nothing wrong, and i looked through the smart report and it was OK.  A few sectors pending allocation but nothing obviously high.  I am not sure if someone might have knocked into the server or not (I am in the process of removing the popcorn/stucco from my ceilings and having plaster put up, so a lot of people are moving around).

 

Anyway... here is order in which things happened:

 

1. parity drive write failed and unRAID kicked it out of the array

2. stopped the array

3. removed the drive on the devices page

4. shut down the server

(the above was all done while at work after i got the email from the server)

5. checked connections for power and data (nothing out of the obvious found to be wrong)

6. started the server back up

7. the array was stopped so i went and reassigned the drive

8. I pressed the start button to bring the array back on line

9. It got about 15% in and I noticed the error count on the Seagate drive start rise, and it got kicked out of the array

10. When i went back to check on the server the seagate was making a clicking type noise so i stopped the server and shut it down.

 

That is where it stands now.

 

The seagate drive is still recognized (for now) and I may be able to start the server back up and pull the data off the drive onto another in the array. I am not sure, but i think i can tell the server that my parity drive is OK (It should be very close to correct as I was not writing much of anything to the array).  I had rtorrent running and it was seeding some stuff but there was no serious writing to the array going on.

 

I am going to try and pull the data off the drive manually first and if that fails try to tell the server that parity is fine and just rebuild the failed Seagate drive.

 

 

Any input would be greatly appreciated.

I'm sure you've already done this, but double check cabling to drives.

 

There are reports of people that put their drives in the freezer and are able to use them long enough to pull off critical data.  I've never had much luck, but YMMV.

 

I had a neighbor that asked for my help.  Turned out he had a crashed hard disk.  All attempts to access it failed.  Clicking like crazy.  No boot.  I put it in an external USB caddy and would not read from my machine.  I figured the drive was dead, but next day he took it into the local mom and pop computer shop (the kind that are disappearing) and they recovered all his data and gave him a boot disk that worked perfect.  I think the service was free (if he bought the replacement drive).  No idea what they did - but if there is critical data on the drive you might consider something similar.

 

Make sure you email Tom.  He may have some ideas but also would be interested to know that this type of failure has occurred.  Your caveat of workman in the vicinity gives some doubt as to whether this event may have been caused by rough treatment, but that seems pretty unlikely.  I'll accept this as a true double disk failure (or if you recover at least a very near miss!)

 

Sorry this has happened to you.  Based on other posts seems like you were doing all the right things from a monitoring and parity check perspective.

 

Best of luck recovering all or most of your data!

 

The freezer trick has worked for me once in the past.  It would make the drive readable for an hour or two, then it would warm up and fail again.  Luckily I only had to save about 60 GBs of data off that drive, so an hour of readable time was enough.  You may be able to refreeze it and try it again, I don't know.

 

If you do try the freezer trick, be sure to place the drive in a ziploc bag first.  If water gets in the drive, it definitely will be dead for good.

 

I really know nothing about drive recovery software for ReiserFS.  GetDataBackNTFS has worked wonders for me in the past, but obviously that won't help here.

 

If this does turn out to be a full-on double drive failure all the way across the sky (so intense), then I guess you can at least count yourself lucky that you lost a parity drive plus a data drive instead of two data drives.

I am in the process of getting some of the data off the drive.  I have it plugged in via USB dock to my test server and am transferring data to a drive in that array.

 

Most of the data was my personal files, that I also have on my local machine, except for some archived stuff.  I think I could have probably gotten that back via crashplan also.  The other stuff I needed to get off of there was an archive of stuff I have downloaded over the years.  Noting on the drive was super critical, but it is still a pain to deal with.

 

 

On a side note I think I am going to take the opportunity to upgrade to 4.5.6 once this is all over and everything is back in the correct place.