May 9, 201412 yr The largest disk in my array was producing many write errors, but unraid did not red ball it. I removed it, checked the cabling and booted the array. It rebuilt the disk. However it was still producing write errors everytime I tried to copy additional data to it. I shutdown the array and low level formatted the drive. It appeared in unraid but, unraid failed to format the disk. The array is not-recreating the files that existed on the drive either, I can only see folders. I have several spare drives that are smaller, that i could put in and I would like to move the data to, IF possible. Otherwise if I have to purchase another disk of the same size I will. What worries me is that the data is not being re-created or simulated. I'm on version 5.0.2 The array is currently started with the drive not connected (physically disconnected) What do you suggest I do? Have I lost the data on that drive
May 9, 201412 yr A low level format probably wiped the drive completely clean. And 'low level format' means different things to different folks. What exact command did you give?
May 9, 201412 yr Author I used a low level format tool on Hiren's boot cd. A maxtor one from memory that is compatible with most drives. (samsung)
May 9, 201412 yr Author I will need to powerdown the array to reconnect the drive to run a smart report. Is that going to make things worse?
May 10, 201412 yr Author I powered down the array and reconnected disk5 with a new sata cable. Powered up and have attached smart report. I have not started the array. Drive looks sick. smart.txt
May 10, 201412 yr Author Shouldnt I be able to see what was originally on disk5 from parity and the remaining 4 disks? I dont recall running a parity check since having problems with the drive.
May 10, 201412 yr Shouldnt I be able to see what was originally on disk5 from parity and the remaining 4 disks? I dont recall running a parity check since having problems with the drive. Your physical drive5 looks like it has failed. It is worthless IMO. The array should be simulating disk5 using parity and the remaining disks. It is simulating something, otherwise would show up as unformatted. When was your successful last parity check? I am unclear of this "low level format" you performed. Did you run it on the physical disk (/dev/sd?), on the array disk (/dev/md5), or on the drive share (//tower/disk5), or another way? I am concerned that whatever you did effectively deleted everything from the simulated disk.
May 10, 201412 yr Author I booted the server off different media (i.e. Unraid was not runing) and formatted the device using Maxtors low level format utility http://www.hiren.info/downloads/freeware-tools/maxtor-low-level-formatter - Not possible for me to have formatted the "simulated" disk5. If i physically remove disk5 now and start the array with the addition of some of these spare new 1tb's drives i have (smaller than the failed disk) will it still try and simulate disk5's data?
May 10, 201412 yr In THIS POST disk5 was being simulated. If you get back to this state it will again be simulated. The array will not start unless either all disks are present, or all disks but one are present (and the missing one is simulated). Are you certain that you did not get a message that a disk was not formatted and select the format option. If this you did that this situation would make more sense and I'd have more confidence the procedure below would help. You could rebuild disk5 onto a fresh disk, and then attempt to recover data from the rebuilt disk using reiserfsck. If you do this, after the rebuild, I would remove disk5 from the array (it is again being simulated), and do the resiserfsck on the rebuilt physical disk (e.g., /dev/sdX1). (Do not forget the 1!). Not only is this faster, it will also preserve the current state of the array and if this doesn't work and suddenly you could try a new idea (nothing occurs to me now but maybe you'll remember something else and another option would be available). Or you could attempt to recover data with reiserfsck directly on the simulated disk. This would be irreversible. You don't need a new physical disk to do this, but since you'll need a new disk anyway, not a big issue. If reiserfsck is able to recover some of the data, you'd need to rebuild parity with the new physical disk5. Someone can talk you through that easily. I'd do the disk rebuild (first option). You will have done no harm and if it doesn't work and you come up with a new idea. you can try it.
May 10, 201412 yr Author it is attepmting to rebuild onto disk5 now. Once completed how do I go about running reiserfsck. (i'm on unraid 5.0.2)
May 10, 201412 yr Rebuilding into a new disk? The smart report posted indicates a bad disk. I would not write anything to it.
May 11, 201412 yr Author Thanks for all your help so far! disk5 has rebuilt - apparently without error. However there is nothing on it. See attached screenshot. Should I run a parity check as indicated in the screenshot or should I stop the array, remove the disk and see what it simulates?
May 11, 201412 yr Thanks for all your help so far! disk5 has rebuilt - apparently without error. However there is nothing on it. See attached screenshot. Should I run a parity check as indicated in the screenshot or should I stop the array, remove the disk and see what it simulates? Not sure you are listening. Your disk5 is bad. The smart report you posted below shows it is bad. As I said before ... Your physical drive5 looks like it has failed. It is worthless IMO. Nowhere did I give advice to rebuild onto that disk. My suggestion was to rebuild on a fresh (i.e., new or otherwise working) disk ... You could rebuild disk5 onto a fresh disk, and then attempt to recover data from the rebuilt disk using reiserfsck. ... If you ask for help in the forums you have to listen. Otherwise you will make bad into worse, and reduce your recovery options. If indeed this reiserfsck recovery that I suggested fails, your best chance of recovery would have been to try to resurrect that disk. Given it can't read it seems unlikely it would have been successful, but now that you have rebuilt on top of it that option is lost. I can't suggest that you attempt recovery on that failed disk, but that is up to you. See http://lime-technology.com/forum/index.php?topic=33071.45 and http://lime-technology.com/forum/index.php?topic=32849.165 for similar situations where users used reiserfsck to do a similar type of recovery. I would suggest rebuilding onto a GOOD disk before attempting recovery with reiserfsck.
May 12, 201412 yr Author Sorry bjp999 in my haste i did not read properly. As I have several 1TB drives spare, I do no wish to purchase a new 2TB just for just this rebuild (if avoidable). I have a spare 3TB external USB drive. I know in older versions this was not supported, can I utilise this external drive for as a member disk temporarily to rebuild onto?
May 12, 201412 yr Sorry bjp999 in my haste i did not read properly. As I have several 1TB drives spare, I do no wish to purchase a new 2TB just for just this rebuild (if avoidable). I have a spare 3TB external USB drive. I know in older versions this was not supported, can I utilise this external drive for as a member disk temporarily to rebuild onto? No. A USB disk can't be part of the array. But you may be able to clone the simulated disk onto the USB disk to run reiserfsck. I do not know how to do this. How important is recovering this data? Seems not very. The measure twice cut once principle should apply. You should be reading and rereading instructions before taking an action. And if this data is important the cost of a new disk would seem trivial. I laid out an option to run reiserfsck on the simulated disk with cons of this approach. Did you consider that?
May 13, 201412 yr Author I laid out an option to run reiserfsck on the simulated disk with cons of this approach. Did you consider that? I would like to try this. What device name should be used and in what state does the array need to be in (maint mode, stopped, started) I am in a state currently where disk5 is being simulated, with the array started.
May 13, 201412 yr I laid out an option to run reiserfsck on the simulated disk with cons of this approach. Did you consider that? I would like to try this. What device name should be used and in what state does the array need to be in (maint mode, stopped, started) I am in a state currently where disk5 is being simulated, with the array started. You should be in maintenance mode and use the dev/md? device name where ? corresponds to the disk number.
May 13, 201412 yr Author I laid out an option to run reiserfsck on the simulated disk with cons of this approach. Did you consider that? I would like to try this. What device name should be used and in what state does the array need to be in (maint mode, stopped, started) I am in a state currently where disk5 is being simulated, with the array started. You should be in maintenance mode and use the dev/md? device name where ? corresponds to the disk number. Ta, just to clarify, I want to run the check on the simulated disk, not the physical one. sdX? sdX1?
May 14, 201412 yr I laid out an option to run reiserfsck on the simulated disk with cons of this approach. Did you consider that? I would like to try this. What device name should be used and in what state does the array need to be in (maint mode, stopped, started) I am in a state currently where disk5 is being simulated, with the array started. You should be in maintenance mode and use the dev/md? device name where ? corresponds to the disk number. Ta, just to clarify, I want to run the check on the simulated disk, not the physical one. sdX? sdX1? You need to start the array in maintenance mode and run the reiserfsck command(s) on /dev/md?, where ? is the drive number (e.g., /dev/md5).
May 14, 201412 yr Author Must have been wiped clean? root@Silo:~# reiserfsck --check /dev/md5 reiserfsck 3.6.21 (2009 www.namesys.com) ************************************************************* ** If you are using the latest reiserfsprogs and it fails ** ** please email bug reports to [email protected], ** ** providing as much information as possible -- your ** ** hardware, kernel, patches, settings, all reiserfsck ** ** messages (including version), the reiserfsck logfile, ** ** check the syslog file for any related information. ** ** If you would like advice on using this program, support ** ** is available for $25 at www.namesys.com/support.html. ** ************************************************************* Will read-only check consistency of the filesystem on /dev/md5 Will put log info to 'stdout' Do you want to run this program?[N/Yes] (note need to type Yes if you do):Yes ########### reiserfsck --check started at Wed May 14 20:15:07 2014 ########### Replaying journal: Done. Reiserfs journal '/dev/md5' in blocks [18..8211]: 0 transactions replayed Checking internal tree.. finished Comparing bitmaps..finished Checking Semantic tree: finished No corruptions found There are on the filesystem: Leaves 1 Internal nodes 0 Directories 2 Other files 0 Data block pointers 0 (0 of them are zero) Safe links 0 ########### reiserfsck finished at Wed May 14 20:18:12 2014 ########### root@Silo:~#
May 14, 201412 yr Looks like you ran a reiserfsck check, not a recovery. Did you actually read the two referenced threads (see reply #17).
Archived
This topic is now archived and is closed to further replies.