March 28, 201511 yr I am trying to change one of my small drives for a larger and all of a sudden started getting errors on a different drive. I recently ran parity checks several times with no issues at all. The syslog is too large, even zipped to attach. I am attaching a link to the syslog download. Here is the link: http://ppl.ug/5zDcvLnZa_E/ By the way, when I try to view the shares, I see none! I am baffled... Update: Now that the "rebuild" has finished, I can see the folders on the apparently new, rebuild drive. However, they are empty. The drive, which is 1.5 TB replaced a 500GB one, yet it still shows the same amount of empty space. The now failed disk 6 shows as empty under the windows browser. I still do not see any user shares. Should I place back the 500GB drive in and try to somehow restore disk6? If yes, how would I do that? I have not rebooted the tower since the rebuild was completed.
March 28, 201511 yr Author I've restarted the tower. Shares now show up. However, there is no change in the disk6 and disk7 status. Updated syslog attached. Any help would be appreciated. Should I place back the disk7, the old good 500GB one and hope to get drive 6 back? syslog-2015-03-28.zip
March 28, 201511 yr Both disk 6 and disk 7 have corrupted file systems and you should run reiserfsck on them. Since you did a parity check and no errors came up, the implication is that those exact same file system errors also exist on the original disk 7. Myself, I would run reiserfsck on the new drives and hang onto the old disk7 just in case. https://lime-technology.com/wiki/index.php/Check_Disk_Filesystems
March 28, 201511 yr Author Ran the reiserfsck --check /dev/md7 and here is the output. reiserfsck --check started at Sat Mar 28 11:31:38 2015 ########### Replaying journal: Trans replayed: mountid 437, transid 35710, desc 5705, len 1, commit 5707, next trans offset 5690 Trans replayed: mountid 437, transid 35711, desc 5708, len 1, commit 5710, next trans offset 5693 Replaying journal: Done. Reiserfs journal '/dev/md7' in blocks [18..8211]: 2 transactions replayed Zero bit found in on-disk bitmap after the last valid bit. Checking internal tree.. \/ 1 (of 4|/ 1 (of 170// 2 (of 86\block 1114113: The level of the node (0) is not correct, (1) expected the problem in the internal node occured (1114113), w/ 2 (of 170|block 1114114: The level of the node (0) is not correct, (2) expected the problem in the internal node occured (1114114), w/ 2 (of 4/block 92770236: The level of the node (0) is not correct, (3) expected the problem in the internal node occured (92770236), whole subtrefinished Comparing bitmaps..vpf-10640: The on-disk and the correct bitmaps differs. Bad nodes were found, Semantic pass skipped 3 found corruptions can be fixed only when running with --rebuild-tree ########### reiserfsck finished at Sat Mar 28 11:32:07 2015
March 28, 201511 yr Ran the reiserfsck --check /dev/md7 and here is the output. reiserfsck --check started at Sat Mar 28 11:31:38 2015 ########### Replaying journal: Trans replayed: mountid 437, transid 35710, desc 5705, len 1, commit 5707, next trans offset 5690 Trans replayed: mountid 437, transid 35711, desc 5708, len 1, commit 5710, next trans offset 5693 Replaying journal: Done. Reiserfs journal '/dev/md7' in blocks [18..8211]: 2 transactions replayed Zero bit found in on-disk bitmap after the last valid bit. Checking internal tree.. \/ 1 (of 4|/ 1 (of 170// 2 (of 86\block 1114113: The level of the node (0) is not correct, (1) expected the problem in the internal node occured (1114113), w/ 2 (of 170|block 1114114: The level of the node (0) is not correct, (2) expected the problem in the internal node occured (1114114), w/ 2 (of 4/block 92770236: The level of the node (0) is not correct, (3) expected the problem in the internal node occured (92770236), whole subtrefinished Comparing bitmaps..vpf-10640: The on-disk and the correct bitmaps differs. Bad nodes were found, Semantic pass skipped 3 found corruptions can be fixed only when running with --rebuild-tree ########### reiserfsck finished at Sat Mar 28 11:32:07 2015 Run it with --rebuild-tree Disk 6 will probably be similar
March 28, 201511 yr Author Looks like it's going to be a while for the --rebuild-tree /dev/md7 to finish. Should I check and then if necessary fix the disk6 in the similar fashion after this one is done, or should I reboot the server and check disk7 first? Thank you, Lev
March 28, 201511 yr Looks like it's going to be a while for the --rebuild-tree /dev/md7 to finish. Should I check and then if necessary fix the disk6 in the similar fashion after this one is done, or should I reboot the server and check disk7 first? Thank you, Lev Really up to you? Personally I am impatient and would reboot the server to see if it appears to have helped.
March 28, 201511 yr Author Really up to you? Personally I am impatient and would reboot the server to see if it appears to have helped. Sounds good. I'll try that.
March 28, 201511 yr Author So, after finishing the rebuilt-tree, I rebooted and now disk7 has two directories, one is "lost+found" and the other "movies" which is empty. I cannot browse the lost+found as I get an error that I do not have permission to view it. Here are the last line of the executed command: Files found: 1 Directories found: 22 Names pointing to nowhere (removed): 54 Pass 3a (looking for lost dir/files): ####### Pass 3a (lost+found pass) ######### Looking for lost directories: Looking for lost files:0 /sec Flushing..finishede 0, 0 /sec Objects without names 4 Files linked to /lost+found 4 Pass 4 - finished done 0, 0 /sec Flushing..finished Any way for me to get those files from the old drive? Should I move on to disk6?
March 28, 201511 yr To get access to the lost+found you should run something like newperms /mnt/disk7 . The reiserfsck would have set the contents as owned by 'root' which is why they are not accessable across the network.
March 29, 201511 yr Author It took a while, but the rebuilt-tree finished. I now have a lost-found folder that I cannot access trough windows. However, I can copy items off of it onto my cache disk through MC. The disk6 is still disabled. What is the best was to recover the files off of it, and subsequently re-enable that disk? Can I somehow re-enable it now and then manually look through the recovered files and restore them under their proper name unto the same disk? Why is that drive still disabled? Thank you. syslog-2015-03-29.zip
March 29, 201511 yr To get the access to lost+found via Windows you need to run newperms as mentioned earlier. A disable disk is only set to no longer be disabled when you rebuild it. Running reiserfsck will have just reset the 'emulated' drive (assuming you were running it against a 'md' type device. If you are happy that the physical disk is OK then the following process will force it to be rebuilt back onto itself: Stop the array Unassign the disk Start the array. It should now be saying it is unprotected and the disk slot in question should say 'missing' Stop the array Assign the disk back to the slot. You should now be told that starting the array will initiate a rebuild of the disk Start the array and the rebuild will start As long as the rebuild completes without error then the 'disabled' status will be cleared.
March 29, 201511 yr Author To get the access to lost+found via Windows you need to run newperms as mentioned earlier. A disable disk is only set to no longer be disabled when you rebuild it. Running reiserfsck will have just reset the 'emulated' drive (assuming you were running it against a 'md' type device. If you are happy that the physical disk is OK then the following process will force it to be rebuilt back onto itself: Stop the array Unassign the disk Start the array. It should now be saying it is unprotected and the disk slot in question should say 'missing' Stop the array Assign the disk back to the slot. You should now be told that starting the array will initiate a rebuild of the disk Start the array and the rebuild will start As long as the rebuild completes without error then the 'disabled' status will be cleared. Will the lost+found directory remain on the rebuild drive? And, this may a be a little far fetched, is there a way for me to place back to old 500GB drive, and copy info of off that one as it appears to have been lost upon upgrading the disk7?
March 29, 201511 yr To get the access to lost+found via Windows you need to run newperms as mentioned earlier. A disable disk is only set to no longer be disabled when you rebuild it. Running reiserfsck will have just reset the 'emulated' drive (assuming you were running it against a 'md' type device. If you are happy that the physical disk is OK then the following process will force it to be rebuilt back onto itself: Stop the array Unassign the disk Start the array. It should now be saying it is unprotected and the disk slot in question should say 'missing' Stop the array Assign the disk back to the slot. You should now be told that starting the array will initiate a rebuild of the disk Start the array and the rebuild will start As long as the rebuild completes without error then the 'disabled' status will be cleared. Will the lost+found directory remain on the rebuild drive? And, this may a be a little far fetched, is there a way for me to place back to old 500GB drive, and copy info of off that one as it appears to have been lost upon upgrading the disk7? Yes, the lost+found directory will be part of the rebuild. Can't quite parse your other question, but you should wait until the rebuild is done, then a parity check to confirm, before you try anything else.
March 29, 201511 yr Author I was replacing a 500gb drive with a 1.5tb one (didk7) when disk6 decided to redball on me. So, I technically lost everything on disk7. However, I still have that 500gb original drive, with the information still on it. Once the rebuild of the now redballed disk6 is completed, can I somehow place the original 500gb disk back in, but not be part of the array? That way I can MC into it and copy all of the original files off of it?
March 30, 201511 yr I don't understand what you did. You cannot rebuild disk6 without disk7 and all the other disks including parity.
March 30, 201511 yr Author I don't understand what you did. You cannot rebuild disk6 without disk7 and all the other disks including parity. OK. Originally, disk7 was a 500GB drive. I took it out of the array and replaced it with a 1.5TB drive, to gain more space. While rebuilding that disk7, disk6 red balled. After the rebuild was completed, disk7 had nothing on it and disk6 was disabled. Sorry if this sounds confusing. That is why I think that the original 500gb disk7 will have its files on it. I was hoping to get those back.
March 30, 201511 yr As you may already understand, red ball means a write to the drive failed and unRAID has disabled it. Normally, unRAID would simulate the drive using the other drives plus parity, but since disk7 was rebuilding I don't know how it would proceed. You say the rebuild completed successfully? Sounds like the original disk7 should be OK, but I'm unclear about the state of the new disk7. Maybe someone else can chime in about that, or with a method to get unRAID to accept your old disk7 so you can rebuild disk6. I have seen similar threads about using invalidslot but I don't have any experience.
March 30, 201511 yr Author As you may already understand, red ball means a write to the drive failed and unRAID has disabled it. Normally, unRAID would simulate the drive using the other drives plus parity, but since disk7 was rebuilding I don't know how it would proceed. You say the rebuild completed successfully? Sounds like the original disk7 should be OK, but I'm unclear about the state of the new disk7. Maybe someone else can chime in about that, or with a method to get unRAID to accept your old disk7 so you can rebuild disk6. I have seen similar threads about using invalidslot but I don't have any experience. Unfortunately, all of the info from the old disk7 was lost during the rebuild, and some of the disk6 files are missing. I was hoping to, without adding the old disk7 back in, to just read off of it.
March 30, 201511 yr As you may already understand, red ball means a write to the drive failed and unRAID has disabled it. Normally, unRAID would simulate the drive using the other drives plus parity, but since disk7 was rebuilding I don't know how it would proceed. You say the rebuild completed successfully? Sounds like the original disk7 should be OK, but I'm unclear about the state of the new disk7. Maybe someone else can chime in about that, or with a method to get unRAID to accept your old disk7 so you can rebuild disk6. I have seen similar threads about using invalidslot but I don't have any experience. Unfortunately, all of the info from the old disk7 was lost during the rebuild, and some of the disk6 files are missing. I was hoping to, without adding the old disk7 back in, to just read off of it. Do you have a spare SATA port so that the old disk can be added outside the array? If not what about a SATA-USB adapter? The vast majority of the time the disk has not physically failed and one can read most of the data off it. If the disk is not mountable then you may need to run a reiserfsck against the old drive first to repair any file system corruption. If so, do not forget that with physical drives the partition number needs to be specified, so the device name is of the form /dev/sdX1 - a common mistake is leaving off the partition number.
March 30, 201511 yr Author As you may already understand, red ball means a write to the drive failed and unRAID has disabled it. Normally, unRAID would simulate the drive using the other drives plus parity, but since disk7 was rebuilding I don't know how it would proceed. You say the rebuild completed successfully? Sounds like the original disk7 should be OK, but I'm unclear about the state of the new disk7. Maybe someone else can chime in about that, or with a method to get unRAID to accept your old disk7 so you can rebuild disk6. I have seen similar threads about using invalidslot but I don't have any experience. Unfortunately, all of the info from the old disk7 was lost during the rebuild, and some of the disk6 files are missing. I was hoping to, without adding the old disk7 back in, to just read off of it. Do you have a spare SATA port so that the old disk can be added outside the array? If not what about a SATA-USB adapter? The vast majority of the time the disk has not physically failed and one can read most of the data off it. If the disk is not mountable then you may need to run a reiserfsck against the old drive first to repair any file system corruption. If so, do not forget that with physical drives the partition number needs to be specified, so the device name is of the form /dev/sdX1 - a common mistake is leaving off the partition number. Yes, the disk is still attached internally to SATA. I have Norco 4220. Can I simply MC into it without enabling it in the array?
March 30, 201511 yr Yes, the disk is still attached internally to SATA. I have Norco 4220. Can I simply MC into it without enabling it in the array? It will need to be mounted first. I would typically use commands of the form: mkdir /mnt/tmp mount /dev/sdX1 /mnt/tmp If those complete without error then the drive will be available at /mnt/tmp.
March 30, 201511 yr Author Yes, the disk is still attached internally to SATA. I have Norco 4220. Can I simply MC into it without enabling it in the array? It will need to be mounted first. I would typically use commands of the form: mkdir /mnt/tmp mount /dev/sdX1 /mnt/tmp If those complete without error then the drive will be available at /mnt/tmp. Thank you very much. I'll give it a shot today.
April 1, 201511 yr Author Yes, the disk is still attached internally to SATA. I have Norco 4220. Can I simply MC into it without enabling it in the array? It will need to be mounted first. I would typically use commands of the form: mkdir /mnt/tmp mount /dev/sdX1 /mnt/tmp If those complete without error then the drive will be available at /mnt/tmp. I finally was able to complete the rebuilt and a clean parity check. I think that two of my WD EARS 2TB drives failed. The rebuild completed successfully, but running parity check generated a lot of errors. Tried a third 3TB seagate and that seemed to complete with no errors. I am copying the contents from the old drive over to the array. Once the copy finishes do I need to unmount the old drive an remove the tmp directory?
Archived
This topic is now archived and is closed to further replies.