Red Dot

[n1]

The power went off yesterday and after I rebooted the server it started a parity check...  I checked this morning and disk 3 had over 500 errors and a red dot (instead of green).

 

I went on to see if I could still stream from it.  I tried maybe ten different videos and they all seemed fine.

 

Should I just replace the disk and rebuild, or should I be doing something else?

 

Any help is greatly appreciated.

[Archivist]

The power went off yesterday and after I rebooted the server it started a parity check...  I checked this morning and disk 3 had over 500 errors and a red dot (instead of green).

 

I went on to see if I could still stream from it.  I tried maybe ten different videos and they all seemed fine.

 

Should I just replace the disk and rebuild, or should I be doing something else?

 

Any help is greatly appreciated.

 

Hi N,

 

I've had a similar problem in the past 2 weeks (red dot on Parity Drive no less), I would recommend a couple of things.  First we need your System Information – unRaid Version (4.2/4.7), Motherboard, Hard Drive (model, type, age), Power supply, etc.  While I think a new Parity calculation is ok, if your drive is failing, that may lead to errors being copied to Parity.  Some of the Guru’s or Hero’s may have some additional thoughts (?).

 

I think we need to see a Smart Report on your Drive and a Syslog for your system.  If you have unMenu installed, you can get them from myMain (“sm”) and Syslog, then save and either zip or post the txt files as attachments.  If you don’t have unMenu installed, try the following:

 

    smartctl -a -d ata /dev/sda | todos >/boot/smart.txt (including spaces)

 

and

 

    cp /var/log/syslog /boot/syslog-2011-07-30.txt

      chmod a-x /boot/syslog-2011-07-30.txt

 

these files will be on your “Flash” drive and you can copy them to your reply post.

 

I am also referring you to the Main Troubleshooting Guide for unRaid for more information:

 

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

 

I’m not a Guru, so I hope some of the Pros will jump in to help, but hopefully that should at least get your started.

 

Good Luck

 

Dave

 

[Joe L.]

The power went off yesterday and after I rebooted the server it started a parity check...  I checked this morning and disk 3 had over 500 errors and a red dot (instead of green).

A disk is marked as disabled (red) if a "write" to it fails.  Apparently, a "write" to your disk3 failed.

I went on to see if I could still stream from it.  I tried maybe ten different videos and they all seemed fine.

That is the beauty of a RAID array.  The contents are currently being re-constructed on the fly from the contents of the parity disk and all the other data disks.  To your media player all looks normal.  You can read the "simulated" disk, and even write to it.  The writes will be made to parity, in fact, you know at least one write was made to it, and that write was un-successful, and the disk disabled.

Should I just replace the disk and rebuild, or should I be doing something else?

 

Any help is greatly appreciated.

You could just replace the disk, but it could also be as simple as a loose connector on the disk, or a bad SATA or POWER cable, or the disk could be loose in a drive tray, or a bad port on the disk controller.

 

The  most common is a loose connector, since SATA connections loosen easily.  Since you had a power outag the temperature in the case might have cooled and shifted things.  it could easily be a loose connector.

 

So...

 

Attempt to get a smart report for the disk.  If it returns one, then you might have an intermittent fault.

If not (and perhaps regardless, since it might be intermittent):

Stop the array

Power down

Re-seat the connectors.

Power up.  

The disk will NOT restore itself.  This is by design.  You must convince unRAID the disk is being replaced.  It will think it is a replacement for itself if you Stop the array, un-assign the failed disk.  Start the array with it un-assigned (this causes the array to forget the serial number of the disk) Then stop the array, re-assign the disk to its slot, and then start it once more.

 

At that point, unRAID will be saying it will re-construct the disk onto its replacement.  Let it.

 

If the disk itself is really bad, then yes, just replace it.  Since it has a different serial number than the existing there is no need to un-assign the old.

 

Joe L.