Tried to upgrade parity to 3TB: FAILED. Now I want to go back

[lovswr]

This is 5.0-rc11

 

Got a new 3TB drive to replace the existing 2TB parity drive.  Did parity check ok.  replaced 3TB  but when I reboot, one of the date drives has yellow ball & now says unformatted.  Put old 2TB parity drive.  Now the parity drive has yellow ball & no option to do parity check & STILL have data drive with yellow ball.  When i browse my shares every thing appears to be ok.

 

Is there some kind of way I can get the old (I think still valid) 2TB parity drive to be recognized, so I can replace/reformat my data drive?

 

 

Here

[Joe L.]

This is 5.0-rc11

 

Got a new 3TB drive to replace the existing 2TB parity drive.  Did parity check ok.  replaced 3TB  but when I reboot, one of the date drives has yellow ball & now says unformatted.  Put old 2TB parity drive.  Now the parity drive has yellow ball & no option to do parity check & STILL have data drive with yellow ball.  When i browse my shares every thing appears to be ok.

 

Is there some kind of way I can get the old (I think still valid) 2TB parity drive to be recognized, so I can replace/reformat my data drive?

 

 

Here

Short answer... IF you saved a copy of the "config" directory BEFORE you swapped in the new parity drive... maybe.  Restore the drives as they were, restore the old config directory contents.  Then reboot.

 

Longer answer...  post a syslog.  post a screen-shot.

I know about red, green, blue, and orange indicators... but not yellow.  Sorry.  It will take some analysis.  Hopefully you've not rebooted and lost the syslog.

 

Joe L.

[lovswr]

No I did not save the config file?

 

Here is the syslog & screen shot.

 

 

edit:  Disk 3 held only my pictures share.  With the array running, that share is no longer there.  If it is really gone (blank or unformatted etc...) I would like to know how I can get it back to a state where I can just add another blank disk to my array.

 

 

 

 

 

 

17_feb_unraid_syslog.txt

[itimpi]

The chances are that the data is intact on your disk, but that a reiserfs error stopped it being mounted which is why it shows up as unformatted.

 

I would suggest that you stop the array; telnet in; and then do a 'reiserfsck /dev/hdd1' command to check the disk filing system.

 

as an aside I notice that you have a couple of disks Including the problem one) showing up as 'hdd' devices.  This normally means that you have them set up in your BIOS as 'IDE' emulation devices rather than AHCI devices.  Changing them to AHCI should make them perform better.

[Joe L.]

It appears as if disk3 is "missing" in your syslog.  You might have dislodged a cable to it in your replacement of the parity drive.

 

Power down, re-seat the cables, then power up.  See if the disk is then detected. 

[lovswr]

Ok I powered down & checked all cables.  As far as I could tell nothing changed.  I ran the reiserfsck command & it aborted. 

 

 

 

[Joe L.]

First, you ran reiserfsck incorrectly.  The file system is on the first partition, not the raw drive.

Second, you would never run it to check a disk itself, you would use

smartctl -a /dev/hdX

or

smartctl -a /dev/sdX

or

hdparm -i /dev/sdX

 

To check the file system on the protected partition you would use the /dev/mdX device.

/dev/md1 = disk1, /dev/md2 = disk2, /dev/md3 = disk3, etc...

reiserfsck --check /dev/md3

 

If the disk does not respond, look in the syslog for its model/serial number.  Its device name may have changed. (they are assigned as disks initialize, so can change fro one boot to the next)

 

If the disk does not respond, it could be a bad cable, or bad drive tray, or backplane, or a bad disk controller port... it is not necessarily the disk that has failed.

 

Joe L.

[lovswr]

Ok I got super-duper lucky.  I had resigned myself to completely losing all data, except for my mail archive which I managed to get off the array.  I ran the initconfig command & then put all the original* disks back in their correct slots.  Array (included over 200G of pictures{whew!}) is back & running parity check now.

 

 

 

 

*the new 3tb is still physically connected but not assigned to the array.  I will definitely let this parity check & then I may even wait a few more days before I try to swap the parity drive again.