These errors happened at the same time. I booted and

 

• Disk 1 has a red ball next to it. Shows "DISK_DSBL" in unmenu
  
• Disk 4 has a red ball next to it. Shows "DISK_WRONG" in unmenu (I did not change any disks)
  
• Disk 4 is showing a size of 2,147,483,616 even though the size is 976,762,552
  
• Disk 4 is also not showing a temperature
  

 

I understand that DISK_DSBL means an error occurred during a write. It seems to me that Disk 4 has failed, not Disk 1. Disk 4 is makes many power-cycling noises and also has the strange problems described above.

 

To verify that one drive wasn't interfering with another, I tried unplugging 1 and leaving 4 plugged in, and vice versa. 1 remained as DISK_DSBL, and 4 remained as DISK_WRONG.

 

Can I FORCE unraid to undo DISK_DSBL on disk 1, or in other words, trust that it's fine? If I can do that, then I can replace disk 4 and rebuild it.

 

syslog:

http://pastehtml.com/view/1cxc42q.html

 

screenshots:

unraid: http://i51.tinypic.com/2f03ggm.jpg

unmenu: http://i52.tinypic.com/2ikbgp.jpg

Your analysis looks spot on!  I think you are right that Disk 1 is completely fine, and Disk 4 is probably toast.  It is operational, and communicating, and the SATA link is working fine, but I would not be surprised if it is not even spinning, or there was a serious head crash.  The drive (Disk 4, the Seagate) reports zero sectors (unRAID seems to have misinterpreted the size, which has really confused it).

 

This syslog is current, well after the disabling occurred, so I cannot say for sure, but I know what usually explains this scenario.  Your onboard drives are set in the BIOS settings to an IDE mode, which is emulated, perhaps too well.  Changing that to AHCI would be better, if only because it avoids one of the flaws of IDE drives, one that bit you.  IDE drives are attached 2 at a time to IDE channels, and unfortunately under some conditions and drive errors, the whole channel is reset or closed, which affects or disables both drives on that channel.  Your Disk 1 and Disk 4 are sharing a pseudo IDE channel.  So when whatever catastrophic issues occurred to Disk 4, the channel was affected and both drives were disabled.  It would have been nice to see the syslog when this happened, to confirm this.

 

Now as you surmised, we've got a problem.  We want to rebuild Disk 4 onto a new replacement drive, but we can't because it thinks 2 drives are down.  I *think* you can use the "mdcmd set invalidslot 4" command to rebuild Disk 4, but I will defer to others to direct you on that.

 

It would be nice to try to get a SMART report for Disk 4 before removing it, to confirm it has failed, but I don't think that is possible, since it cannot even report a temperature.

Thanks for the info Rob. Disk 4 is definitely messed up. The bios loading screen showed the error "4th Slave Hard Disk Error" and I can't run a smart report. Smart report on disk 1 is fine.

 

Can anyone tell me how to force it to re-enable disk 1? If I can't do this, I can't rebuild disk 4. Nightmare situation.

I am not expert at all, but what if you replace the bad drive with a good one then let it rebuild?

I am not expert at all, but what if you replace the bad drive with a good one then let it rebuild?

because there are TWO disks disabled, he cannot start the server to let it rebuild.

 

The series of commands would be

log in and type

initconfig

Respond to its prompt with "Yes"  (Capital "Y", lower case "es")

Then, before doing anything else, and before starting the array type

mdcmd set invalidslot 4

Then refresh the web-console.  It should show all drives as "blue" and you can press "Start"

When you start the array it will attempt to rebuild disk4.  (If you had a replacement for the failed drive you could install it in place of the bad disk before typing "initconfig"

 

Joe L.