Here what happened..

 

logged into my tower to find my disk 3 disabled due to write errors

 

I was like okay.. no prob.. must be a problem disk I'll simple read the manual to replace the bad disk.

 

quote from manual : Replace a failed disk

 

This is the case where you have replaced a failed disk with a new disk:

 

Stop the array.

Power down the unit.

Replace the failed hard disk with a new one.

Power up the unit.

Start the array.

 

sounds easy right:  so I hit stop array.... when I refresh, a bunch of my disks are now showing Unformatted !!!

 

what do i do now ?!?!?  should I hard shut down, replece the bad disk and hope for the best?  how can i fix this?  is my data recoverable?? whats wrong?

 

 

 

First step: Do Not Panic!  There is almost certainly nothing lost, unless you move too quickly and take the wrong action here.

 

From the FAQ:  Why is a disk showing as Unformatted?  It begins with the all-important advice to NOT format anything, and do NOT Start the array.

 

Those instructions you found are out-of-date, unfortunately.  Very often, disk errors may not actually involve the drive itself, they may actually be cable or power related.  So the instructions SHOULD say, get a SMART report and capture the syslog and either analyze it yourself, or make it available for others to look at.  Unfortunately, this has to be done BEFORE you reboot, because the syslog is kept only in memory, and lost with each reboot.  Still, if you can provide a syslog now, it may show what kind of errors you are really getting, on Disk 3 (if it is still installed).

 

You also need to get a SMART report for Disk 3, see the Troubleshooting page, Obtaining a SMART report section.

 

Right now, powering off, then back on, will clear the issue with drives appearing to need formatting.  Please re-install the original Disk 3 if you have removed it, then boot and get that SMART report for the drive, and capture the syslog (instructions also on the Troubleshooting wiki page), and post both reports here, and we can advise you further.

Do not panic.

 

Part of the shutdown process of the array is to un-mount all the disks.  All of those currently showing (incorrectly) as un-formatted are simply un-mounted.

 

The one disk not showing as un-formatted was not able to be un-mounted because it was busy. As a result, the array was not able to be stopped.   Unfortunately, it does not re-mount the drives it un-mounted, and the web-interface sees no file-system on those mount points, and then incorrectly shows their status as un-formatted instead of un-mounted.

 

A disk is busy if one of the following is true:

A file on it is currently open for reading or writing

A file on it is currently being invoked (as a program)

A folder on it is the "current directory" for some process.  (You logged in a cd'd there, or "cd'd there and started a process)

 

To proceed, log off, or cd to / if you are logged in via telnet.

Or stop the process you started

Or if it was something that has already stopped, press

"Stop" once more...  If the array can un-mount all the drives, it will un-mount the last drive.

 

If not, you can type:

fuser -km /mnt/disk4

umount /mnt/disk4

and then press stop once more.  (The fuser command will kill any processes keeping the disk busy)

 

This is a known "bug" and is being addressed in future versions as already discussed in this thread: http://lime-technology.com/forum/index.php?topic=3666.msg32835#msg32835

 

Whatever you do, do NOT click on the "format" button.   It would do as you ask... and erase your data.

See this entry in the wiki: http://lime-technology.com/wiki/index.php/FAQ#Why_is_a_disk_showing_as_Unformatted.3F

 

Joe L.

okay; I just refreshed the web management again;

 

now.. the array is stopped; i have the ability to power down (havent yet); and disk3 is not showing "Not Installed"

 

 

- I'm currently reading the troubleshooting to figure out how to do a syslog & smart report % and such; keep in mind I've never telnet'd before, but it doesnt look to hard to learn.  If I do have to be at the local machine, I will have to do it tomorow. 

 

Leaving this running/powered on in this state isnt going to hurt things right?

 

.. also at the moment I'm at work (and will be for the next 12hr) .. I'm curenttly remote desktopping into a pc at home; from her/there I will have to manage things.

 

I'll keep updating this thread; any more info would be great

 

thanks so far

You can do this via remote desktop

From a command window type

telnet tower

you will get a login prompt

the login ID is

root

If it asks for a password, just press enter (unless you set a root password)

Once you log in, type

cp /var/log/syslog /boot/syslog.txt

 

Once you do that, you can get access to the file via windows file explorer at

\\tower\flash\syslog.txt

 

Zip it up, post it here as an attachment.

 

You error can be as simple as a loose connector to the drive, or a bad cable, or bad power connection/splitter.  It is not necessarily a defective drive.

 

Joe L.

here is the syslog (i snipped out some, due to huge file.. also some of the files logged are of the xxx nature, lol)

 

so...

 

 

The small piece you posted had none of the errors... and is almost useless for any analysis.  It was actually very small... some syslogs are many megabytes when un-zipped.  My syslog currently is over 1000 lines long... and I'm not doing too much logging of extra stuff.

 

You can delete the lines with specific file names....  We don't need to know your specific taste in "movies" 

 

The other lines  in the syslog you did not include are very important though for any analysis, unless you really don't think you need any help in interpreting the problems with your array.

Was that the very first error reported for ata18 (sde)?  It does look to me like a connection issue, NOT a problem with the drive, but we really need to see the SMART report, for the health of the drive.

Jun 11 13:12:46 DV_Server kernel: ata18.00: disabled

Once it is disabled, I don't trust any errors that appear *afterward*.

yeah rob, that is the first reported error in the syslog.  So your saying connection issue, ok, well when i get home as my night shift I will verfiy to see if thats the case.  Also I am using a norco 4020 with backplate and all; perhaps I'll also connect the drive directly as to minimize any other variables.

things are normal again; replace the "bad" wd green 1.5 with a spare seagate 1.5 I had laying around. 

 

Unraid worked fine, and rebuild the drive/data

 

back to normal now; the WD drive is new, so I'll simply do a RMA with ncix.

 

thanks for the help; learned alot about syslog's and such

things are normal again; replace the "bad" wd green 1.5 with a spare seagate 1.5 I had laying around. 

 

Unraid worked fine, and rebuild the drive/data

 

back to normal now; the WD drive is new, so I'll simply do a RMA with ncix.

 

thanks for the help; learned alot about syslog's and such

Did you do a "smartctl" report on the drive? or is it really dead and unresponsive?

 

Glad everything worked smoothly.

 

Joe L.

there is definitely something wrong with the drive, tried a reformat in windows, and it freezes part way through and clicks.

 

no big deal, thats what warrenty is for.

Yup, and unRAID protected your data... That is what it is for. 

 

Joe L.