imagnome Posted August 8, 2013 Share Posted August 8, 2013 To make I long story short... my system is running 5rc16c and has a 7 drive array and 1 parity. After an unexpected power-loss I believe that my parity is no longer valid and disk 5 is corrupted. On disk 5, reiserfsck --check is giving a rebuild-sb suggestion. All other drives are fine according to reiserfsck. I rebuilt disk 5 twice with different drives/sata cables and got the same results which leads me to believe that my parity might be faulty. The data on disk 5 is backed up to an external drive. I think my next move should be to remove disk 5 altogether and create an array from all of the other drives and then rebuild the parity. I read that to accomplish this I should do the following: Stop the array by pressing "Stop" on the management interface. Un-assign the drive on the Devices page, then return to the unRAID Main page. Select the 'Utils' tab Choose "New Config" Agree and create a new config If I do this, I won't loose the data on all of the remaining drives, right? I will then rebuild the parity and once again have a protected array, without disk 5. I would really appreciate some confirmation before I take this step. Thanks in advance. My full description and syslog are here: http://lime-technology.com/forum/index.php?topic=28797.0 Quote Link to comment
Joe L. Posted August 8, 2013 Share Posted August 8, 2013 To make I long story short... my system is running 5rc16c and has a 7 drive array and 1 parity. After an unexpected power-loss I believe that my parity is no longer valid and disk 5 is corrupted. On disk 5, reiserfsck --check is giving a rebuild-sb suggestion. All other drives are fine according to reiserfsck. I rebuilt disk 5 twice with different drives/sata cables and got the same results which leads me to believe that my parity might be faulty. The data on disk 5 is backed up to an external drive. I think my next move should be to remove disk 5 altogether and create an array from all of the other drives and then rebuild the parity. I read that to accomplish this I should do the following: Stop the array by pressing "Stop" on the management interface. Un-assign the drive on the Devices page, then return to the unRAID Main page. Select the 'Utils' tab Choose "New Config" Agree and create a new config If I do this, I won't loose the data on all of the remaining drives, right? I will then rebuild the parity and once again have a protected array, without disk 5. I would really appreciate some confirmation before I take this step. Thanks in advance. My full description and syslog are here: http://lime-technology.com/forum/index.php?topic=28797.0 sounds like you are not running the correct reiserfsck command. Please post the exact full command you attempted that suggested rebuid-sb as the next step. Quote Link to comment
garycase Posted August 8, 2013 Share Posted August 8, 2013 No, you won't lose any data on the other drives with a New Config. Simply Stop the array; go to the Utils tab, and click on New Config ... you'll have to confirm this is what you want to do. Then go to Main and assign the disks you want (all of your data disks except the defective one). The only way you could lose data is if you assign the wrong disk as parity => this would then write the parity information to that drive (overwriting all other data on that drive). So be CERTAIN you know which drive is parity. Quote Link to comment
garycase Posted August 8, 2013 Share Posted August 8, 2013 Based on Joe's comment, you may not need to do the new config => if you've in fact not run the correct command, the disk may be rebuildable using the correct reiserfsck commands. Personally, since you have all the data backed up, I'd be tempted to just rebuild the array anyway ... but it certainly wouldn't hurt to do the correct rebuild => and then just verify the data against your backups. Quote Link to comment
imagnome Posted August 8, 2013 Author Share Posted August 8, 2013 Wow, thanks for the fast replies. I started in maintenance mode and ran this command: reiserfsck --check /dev/md5 and it recommended rebuid-sb. I ran the same command on md1, md2, md3, md4, md6, and md7 and upon finishing the scan it said "no corruptions found" on each. I ran the same command on md5 after rebuilding it twice with different drives/cables and it still recommended rebuild-sb. I just transported my system to a different town so I am going to check 1-4,6,7 once again to verify that they are good before I proceed to rebuilding the array without disk 5. All of this is time permitting as my life is pretty much just work and sleep Thanks confidence moving forward. Quote Link to comment
snowboardjoe Posted August 10, 2013 Share Posted August 10, 2013 On a similar note, I had an issue with adding a new cache drive last week, but not using it yet. When I cam back to my config this morning to add two more data drives to my array (after I recently pre_cleared them), I discovered that I assigned my previous drive as a data drive instead of a cache drive--grrrr. So, following some directions, I ran initconfig from CLI, reassigned all of the drives (including the new one) and started the array. Fortunately everything had simple labels so it made it easy to reassign everything. Parity check is zipping along now and will be done in a few hours. However, it still scares me what might happen if I ever screw that up later on. What happens if you swap a data drive slot or something else like that? Can unRAID tolerate data disks being shuffled around? Quote Link to comment
dgaschk Posted August 11, 2013 Share Posted August 11, 2013 The ordering of the data disks makes no difference. Nor does which physical connection used. The only disk which must be correct is the parity drive (and cache drive). Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.