did my "new" drive just die??

[Brydezen]

Hello guys,

 

Just got some bad news on my Prowl app, as I was watching tv. One of my disks - the first one I got, besides the parity disk, the disk that still holds almost ALL my data just got an read error on it. Now there is a red X and the text "Device disabled, content emulated" when I hold the mouse over. I just started a parity check, I don't know if that will fix anything - and I have to sleep right now, school tomorrow sadly. Hope it wasn't a bad decision to start a parity check. And hope to hear from someone real soon hopefully before I go to bed :-( 

 

EDIT: btw it's disk 1 that says it got an read error.

 

Here are my diagnostics. Best regards, a sad "new" unraid user :-(  

tower-diagnostics-20180206-0206.zip

Edited February 6, 2018 by Brydezen

[Frank1940]

First, Don't panic!!!  Although it appears that you are already in a panic mode.  Stop, take a few deep breaths, ...and wait until some one with more experience than I have can get back to you.  (Starting that parity check was not the best idea...)  

[SSD]

Was that diagnostic file taken before a reboot? If so, reboot and take another one and post it.

 

The most common reason for a drive dropping offline is a bad or loose cable. But almost as common these days is using a Marvell based controller like a Supermicro HBA.

 

Chances are extremely good the drive is fine. A failing drive does not typically present with a write error (drives are kicked for write errors or if the drive drops offline, which is what I am expecting happened here). Rebooting and pulling a new diagnostic will typically allow the drive to be visible again and give info to see if the disk is still healthy.

 

Even if the disk did fail, parity and the other disks are emulating that disk. And that means no data loss yet.

 

Wait for someone to report that you are not using a Marvell controller (you can search the syslog for "mvsas" to rule that out of you're not sure). Or see any other problems in your syslog. If not you need to shut down and resecure the data cable.

 

But even if the disk comes back, you still have to repair the array. Nothing you do will put humpty dumpty.back together again by itself. You'll need to rebuild the failed disk to render the most up to date data on the drive. You could rebuild onto the old disk, but I do not recommend it. If it doesn't go well, you would have destroyed your plan B. Therefore I suggest rebuilding into a different disk.

 

But I'm getting ahead of myself. Reboot and post your second diagnostic file and await further feedback.

 

If you don't have a disk large enough to rebuild on, you might look for a good priced option in case you need it.

 

[Brydezen]

I tried to reboot in the hopes that, it had just gone offline - but not it says the drive is not installed :-( I know it not a Marvell controller issue, my motherboard has some of the controllers, but is not used for any of my drives, as I ran into issues with it on the second day of using unraid

 

but anyways here are the new diagnostics after a reboot :-)

tower-diagnostics-20180206-0929.zip

 

EDIT: How can I test what drive is installed here? I may be thinking about labeling them for "easy" replacement of cables or so on for later. Would like to first of all maybe replace the SATA cable just to see if that is the issue - i just really hope no other of my drives is going to fail soo. Don't wanna read or write anything to my array right now

Edited February 6, 2018 by Brydezen

[JorgeB]

Disk is missing after the reboot, and when it dropped offline looks like the disk lost power (or died), so check/replace power cable and try again.

[Brydezen]

I have the same of power cable for all my hard drives? And as I dont know what drive is what. then I don't know what drive to try and replace the power cable or sata cable for :-(

 

EDIT: Could I maybe try and spin down the other drives, to see what drive is installed where in my mount brackets?

Edited February 6, 2018 by Brydezen

[JorgeB]

It's this one:

 

ST4000DM004-2CV104 s/n ZFN06CX6

[Brydezen]

I found the drive, didn't really think of that it would be labeled (stupid me) but I followed the SATA cable, and it still sits in tight like all the other disks. And when I lay my hand on it, sure feels like it is still spinning - so it must get power? So maybe its the SATA cable that went bad? Or do u recommend to also change the power cable? The drive is between two other drives. My parity drive is the one above it, and Disk 2 is the one underneath it

[JorgeB]

Swap cables with another drive just to rule them out.

[Brydezen]

But is that safe? Don't unraid complain if I just switch SATA cables from another drive? When I maybe will miss two drives in the array? And all my data is lost?  I can take the harddrive out, and out it on the same power cable as my ssd, it got it's own cable. And got some spare sata cables laying around, is it not safer to just use a "new" sata cable? 

 

Sorry for all the questions, you are a champ for helping me out here! I'm pretty lost, and sad about this drive failing on me :-(

[JorgeB]

Disks are tracked by serial number, it's fine to change ports or controller

[Brydezen]

So u are just suggesting to use the same power and sata cable, but from a different port on the mb? And is it safe, if i take one from a working one and just swap these two when i shutdown and power up the server again? Isn't it bad if the array suddenly thinks two drives are missing?  

[JorgeB]

3 minutes ago, Brydezen said:

Isn't it bad if the array suddenly thinks two drives are missing?  

No, it just won't start.

[Brydezen]

I just changed the sata cable to a different one, but now it still says the disk is not installed, but it shows up under unassigned devices, with the same drive letter sde as it had in the array. U even used the same port it was in before

New diagnostics uploaded for latest boot. 

tower-diagnostics-20180206-1224.zip

[JorgeB]

SMART looks fine, you'll need to rebuild to old disk or to a new one if you to play it safer.

[Brydezen]

If the drive is fine, but was just the SATA cable messing with me, then I will just rebuild to the same disk. But how do I do it? And do you think it's a bad idea? I got a 2.5" harddrive from my laptop, i'm not really using anymore, but its only 1TB compared to this 3.5" that is 4TB :-( What do you think would be the best solution, and what solutions do I got?

[JorgeB]

4 minutes ago, Brydezen said:

But how do I do it?

http://lime-technology.com/wiki/Troubleshooting#Re-enable_the_drive

 

5 minutes ago, Brydezen said:

And do you think it's a bad idea?

Usually it's fine, but if something goes wrong during the rebuild it can leave the old disk corrupt, so it's always safer to use a spare if available, but it also depends how important is the data and if you backups, and if you don't have backups should be because it's not that important.

[Brydezen]

Hmmmm, I could live with the drive failing doing rebuild, the data is kinda important, as it has all my games, movies and ISOs, but its nothing i cant "rebuild" over time, by redownloading it all again. So I think I'm just gonna rebuild the drive as is

 

EDIT: I did as provided in the re-enable the drive link, and it says the parity / drive is being reconstructed - thank you SOOOO much for the help! But danm 16 hours to rebuild, that is a looong time just because of a bad SATA cable

 

but if U have the time for it, would u not mind taking a look at my Disk 3 (ST31500S41AS_6XW0BBNY) Its an oooold drive from an external drive enclosure where the PCB was bad, and non function for the drive. But it works now in my array, precleared and tested it a bit before adding it. But some of the SMART indicators does "light up" Like the Reallocated sector count and Reported uncorrect. And again thank U so much! 

Edited February 6, 2018 by Brydezen