Joe L. Posted August 19, 2011 Share Posted August 19, 2011 Done this 3 times now. Each time it gives me an error when I run rebuild command. Do you know much about the -B command so I can fix the bad block? It's now showing as O.K. on the web browser, but you can see when it was formatting on the console it was having issues agailne with locked inode and i/o failure. Any ideas? I don't know anything about the "-B" option. sorry. I'll do a google search... see if I learn something. Joe L. Link to comment
Roger Posted August 19, 2011 Author Share Posted August 19, 2011 While I'm waiting, I figure I'll try a rebuild on /dev/md7, seems to be running but just started and it's only been going 1/2 hour. Not sure what to do with Disk 8. From what I'm reading it seems like some kind of block fix using the -B command in reischk, but I don't know how to use that without help. So,I'm pretty much stuck, afraid to start the parity without getting Disk 8 taken care of. Not sure if I have another 1.5 tet around, but if I loaded that instead it would be empty. Link to comment
Roger Posted August 19, 2011 Author Share Posted August 19, 2011 Obviously, without your help, I don't know anything about anything and certainly not the -B option, but; it is the message I get when the rebuild fails on disk 8 Link to comment
Roger Posted August 19, 2011 Author Share Posted August 19, 2011 I have a bunch of extra 1 terabyte drives around. I know Disk 8 used to be a 1.5 tera, and it used to be pretty close to full, or it still is, just can't seem to get it to work. If I start parity, it will try and write to it, correct? Then it will likely fail, because of the bad block, correct? So, what ever this fix the bad block message with -B is, what would your normal move be if you have a disk in my situation that is bad, and your rebuilding a system? Would you fix all other disks first before starting parity? then replace Disk 8? If so, could I replace it with a small drive? Was hoping to get this fixed so I could run the rebuild on 6 disks tonight. Also, is it F3 to go to the third screen, F4 to the 4th until you get to 6? Roger Link to comment
Roger Posted August 19, 2011 Author Share Posted August 19, 2011 Joe L, this is what I found, and frankly this is so far beyond anything I could learn or try myself, I don't think it's a viable solution. So, again I ask you , if you are lame like me, what do you try next? http://smartmontools.sourceforge.net/badblockhowto.html#reiserfs_ex Apologize for all the questions, but; I'm wondering if, before I run the rebuild command on the other disks, after I fix 8 and of course, after Disk 7 is complete, because I can't stop it. I try and copy off everything to a backup. I noticed in Disk5, which you said I could write to, because I ran a rebuild on it, everything being in the lost+found folder, that I found a lot of movies perfectly intact. I also found dozens of movies, perfectly intact but without the name of the directory, they would normally be in. So I know it's going to take me hours to open each one, make the correct directory, and copy the file to it, might use the shares when everything is up to do that. Worried, that when I run rebuild on the other disks, the same thing will happen, when it looks like I could just move everything now. Am I missing something here? Roger Link to comment
Joe L. Posted August 19, 2011 Share Posted August 19, 2011 Joe L, this is what I found, and frankly this is so far beyond anything I could learn or try myself, I don't think it's a viable solution. So, again I ask you , if you are lame like me, what do you try next? http://smartmontools.sourceforge.net/badblockhowto.html#reiserfs_ex Apologize for all the questions, but; I'm wondering if, before I run the rebuild command on the other disks, after I fix 8 and of course, after Disk 7 is complete, because I can't stop it. I try and copy off everything to a backup. I noticed in Disk5, which you said I could write to, because I ran a rebuild on it, everything being in the lost+found folder, that I found a lot of movies perfectly intact. I also found dozens of movies, perfectly intact but without the name of the directory, they would normally be in. So I know it's going to take me hours to open each one, make the correct directory, and copy the file to it, might use the shares when everything is up to do that. Worried, that when I run rebuild on the other disks, the same thing will happen, when it looks like I could just move everything now. Am I missing something here? Roger What I found was this page... http://www.linuxquestions.org/questions/linux-hardware-18/bad-blocks-on-reiserfs-522717/ Translating it to the unRAID environment, (where we have a flash drive, but no floppy disk) it then becomes this series of steps: badblocks -v -o /boot/badblocks_disk8 -b 4096 /dev/sdX1 (where sdX1 = the four character device name of the disk8 reiserfs partition. ) Then reiserfsck --badblocks /boot/badblocks_disk8 /dev/sdX1 (again, /dev/sdX1 = the first partition, where the file-system resides) followed by: reiserfsck --rebuild-tree -S --badblocks /boot/badblocks_disk8 /dev/sdX1 Link to comment
Roger Posted August 19, 2011 Author Share Posted August 19, 2011 Since Disk 8 is now assigned, do I need to unassign it first to do that series of commands? Or just use the umount first and the dev/md like I"m doing with the other drives? Also, what it this doesn't work? I know Disk8 is actually the disk that was being formatted when I shut the machine off, so it probably is just a couple bad block damaged by that. I"m just wondering why I can't just put in a new disk once I've rebuilt the raid and all other disks have been rebuilt, or just replace the drive. Hate to keep bothering you, but; wondering about my other questions too. Really above my head here? Also, doesn't seem like I can run rebuild on more than one disk until I get parity set, is that correct? Link to comment
Joe L. Posted August 19, 2011 Share Posted August 19, 2011 Is disk8 the new disk? or did it have data on it previously? Link to comment
Roger Posted August 19, 2011 Author Share Posted August 19, 2011 Joe L, thanks for your continued help. Disk 8 is a disk that not only had data, it was pretty close to full. 1.5 tera and more than 1.4 tera used. As I got a little smarter, I learned it was a good idea to leave more room on the disks, but by Disk 8, hadn't figured that out yet. The new disk is disk 12, and it is virtually empty. I copied two moves on it, before you reminded me how stupid I was. So, my big question for you, is if you are me, what do you do? Do you continue to try and save Disk8, and if so, in order to run your command, do I need to unassign it, or can I just umount and than run commands as /dev/md8 Link to comment
Roger Posted August 19, 2011 Author Share Posted August 19, 2011 Joe L, stuck in the mud, until you tell me what to do. You got me this far. Link to comment
Joe L. Posted August 19, 2011 Share Posted August 19, 2011 Joe L, thanks for your continued help. Disk 8 is a disk that not only had data, it was pretty close to full. 1.5 tera and more than 1.4 tera used. As I got a little smarter, I learned it was a good idea to leave more room on the disks, but by Disk 8, hadn't figured that out yet. The new disk is disk 12, and it is virtually empty. I copied two moves on it, before you reminded me how stupid I was. So, my big question for you, is if you are me, what do you do? Do you continue to try and save Disk8, and if so, in order to run your command, do I need to unassign it, or can I just umount and than run commands as /dev/md8 I honestly think it would work with either /dev/md8 or /dev/sdX1 (I forget which device it was) To be sure, and since it was nearly full, I'd stop the array and do it on the /dev/sdX1 device. ( I don't think you need to un-assign it as long as the array is stopped) The badblocks command will map the bad blocks, and should let you fix the file system. To do it through the /dev/md device would be slower, that's for sure. You don't need that right now, since you do not have parity established, so use the /dev/sdX1 device. You are getting close, and we are trying to recover as much as possible. no sense in giving in now. Joe L. Link to comment
Roger Posted August 19, 2011 Author Share Posted August 19, 2011 O.K., thanks I'll give it a whirl. Just in case you want to enjoy your weekend. What do I do if it doesn't work? Roger Link to comment
Roger Posted August 20, 2011 Author Share Posted August 20, 2011 Joe L, Don't know exactly how this is going to turn out, but; I am very grateful for you help. I have run the rebuild on Disk1, Disk 2, Disk5, Disk7, and Disk8 is clogging along. Pretty close to done with the first part of the badblocks commands. I'm also simultaneously running on Disk12 the rebuild. Interestingly, with virtually nothing on it, it comes up with way more information to check. Other 1.5 tera's come out checking 244190637, and it came up checking 366265249. I'm a little scared to run the check on Disks9,10, 11, 13, and 14, because most of my new movies are on them, and one thing that seems to happen is if lost+found is created, it loses directory tiers, which are shared, so it's hard for me to know whether the go into my New_Movies directory or my Older_Movies directory, or TV, etc. Some of the disks are included in 5 or 6 different shares. The earlier disks typically have only 2 or 3 shares. Anyway, just wanted to give my thanks, once more, and give you an update. Roger Link to comment
Roger Posted August 20, 2011 Author Share Posted August 20, 2011 Joe L, first command worked perfect and found 19 bad blocks. Command: badblocks -v -o /boot/badblocks_disk8 -b 4096 /dev/sdo1 Second command: reiserfsck --badblocks /boot/badblocks_disk8 /dev/sdo1 failed. It just gave me the whole list of reiserfsck command from check, to clean, and the options, and expert options. Also interesting. When I start array, I can actually see one directory, Older_Movies, and one movie in it. Tempted to just run the third command, but; not without your advice. Any ideas? Disk8_Bad_Blocks.txt Link to comment
Roger Posted August 20, 2011 Author Share Posted August 20, 2011 Moving 500 movies from a lost+found directory with a number to one with a name. Yuck, but; they are there Link to comment
Joe L. Posted August 20, 2011 Share Posted August 20, 2011 Joe L, first command worked perfect and found 19 bad blocks. Command: badblocks -v -o /boot/badblocks_disk8 -b 4096 /dev/sdo1 Second command: reiserfsck --badblocks /boot/badblocks_disk8 /dev/sdo1 failed. It just gave me the whole list of reiserfsck command from check, to clean, and the options, and expert options. Also interesting. When I start array, I can actually see one directory, Older_Movies, and one movie in it. Tempted to just run the third command, but; not without your advice. Any ideas? I would try the third... Joe L. Link to comment
Roger Posted August 20, 2011 Author Share Posted August 20, 2011 Maybe I'm getting a little smarter, I was going to try it, but; wanted to wait. After I save 8, is that when it's safe to set parity, and then try and build 6 drives at the same time, F1,F2, F3 and so on? I don't want to getting too many going at the same time, because of the directory renaming and moving issue. Some come out totally clean. Disk3 had virtually nothing that was an issue. Disk 5, had all the movies in numbered directories, but because of My Movies XLM's, I knew what all but a few were, could rename directory and move. Not sure if My Movies will be able to find without a remap, but; have to be done to try that. Roger Link to comment
Roger Posted August 20, 2011 Author Share Posted August 20, 2011 Joe L, when I'm done with the 3rd command for Disk8, which it did take, do I run anything else on Disk 8? Or just start the array, and if my files are in there hit parity, then fix the other disks? Also, wondering if there is a benefit or detriment to starting the rebuild command on the other 7 disks, not yet done via the console using /dev/sdx rather than waiting to start the array and using the md command, or vis a versa? Roger Link to comment
Joe L. Posted August 20, 2011 Share Posted August 20, 2011 Joe L, when I'm done with the 3rd command for Disk8, which it did take, do I run anything else on Disk 8? Or just start the array, and if my files are in there hit parity, then fix the other disks? I would start the array. Also, wondering if there is a benefit or detriment to starting the rebuild command on the other 7 disks, not yet done via the console using /dev/sdx rather than waiting to start the array and using the md command, or vis a versa? Roger Either way will work. It will go faster fixing the disks without parity. For that reason it might be best if you did all the rebuilds before you assigned parity. It sounds like you are getting there. Link to comment
Roger Posted August 20, 2011 Author Share Posted August 20, 2011 Bummer, on Disk 8 exactly where it alays fails, aboiut 192,133,767 left to go, Cannot read the block 52051000, which I think is the first one that shows up on the report I loaded yesterday. ideas? Roger Link to comment
Roger Posted August 20, 2011 Author Share Posted August 20, 2011 Trying to get 3 working at the same time in console, 2 works fine, gets to the third, freezes on the third. Lost the web screen as well, Decided to Control C out of the 3rd rebuild, but main.htm on the web //tower/ still not coming up. Anything to worry about, or just complete rebuild on the two disks that are running, and then try and address Disk8 again, with hopeuflly another idea? Roger Link to comment
Joe L. Posted August 20, 2011 Share Posted August 20, 2011 Bummer, on Disk 8 exactly where it alays fails, aboiut 192,133,767 left to go, Cannot read the block 52051000, which I think is the first one that shows up on the report I loaded yesterday. ideas? Roger Nope... need to read up on badblocks and how to deal with its output. Probably can write to that one block and then if the disk re-allocates it, you'll be able to read it. Link to comment
Roger Posted August 20, 2011 Author Share Posted August 20, 2011 O.K., I'll stay tuned. Any ideas why I can rebuild on more than 2 drives at a time, or why the web interface fails. I' O.K., once the two drives are done, correct? Get a waiting for http:// message then it fails. I have my server with a fixed i.p. 198.168.1.50. Don't know if it's better for it to find it automatically or not, but; when I was having trouble with my other unit, I went to fixed I.P. :'( Roger Link to comment
Joe L. Posted August 20, 2011 Share Posted August 20, 2011 O.K., I'll stay tuned. Any ideas why I can rebuild on more than 2 drives at a time, or why the web interface fails. I' O.K., once the two drives are done, correct? Get a waiting for http:// message then it fails. I have my server with a fixed i.p. 198.168.1.50. Don't know if it's better for it to find it automatically or not, but; when I was having trouble with my other unit, I went to fixed I.P. :'( Roger easy, not enough ram. Link to comment
Roger Posted August 20, 2011 Author Share Posted August 20, 2011 Such web menu will come up once the two rebuilds are done? Do you use fixed IP or not on server? Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.