June 1, 200917 yr Ok, I was told it was one in a million, but I guess it happened to me. I've had 2 data disks fail at the same time and now am left with 3 good data disks and 1 good parity disk. Disk1 and Disk4 now show up as missing and unRAID is waiting for input from me before knowing what to do. I guess assuming my disks are completely shot, I need to rebuild the array based on just disk2, disk3, and disk5. Do I just do a restore? I want to make sure that this just builds new parity without destroying the data that's left on the good drives. Also, can I leave the good drives connected to the same SATA ports so they continue to show up as Disk2, Disk3 and Disk5 without a Disk1 or Disk4 in the system? I'll replace the 2 bad drives after I've done some additional debug to determine they are in fact bad. Also, any advice on seeing if my data is still good on the bad drives is appreciated. Thanks, Mark
June 1, 200917 yr Ok, I was told it was one in a million, but I guess it happened to me. I've had 2 data disks fail at the same time and now am left with 3 good data disks and 1 good parity disk. Disk1 and Disk4 now show up as missing and unRAID is waiting for input from me before knowing what to do. I guess assuming my disks are completely shot, I need to rebuild the array based on just disk2, disk3, and disk5. Do I just do a restore? I want to make sure that this just builds new parity without destroying the data that's left on the good drives. Also, can I leave the good drives connected to the same SATA ports so they continue to show up as Disk2, Disk3 and Disk5 without a Disk1 or Disk4 in the system? I'll replace the 2 bad drives after I've done some additional debug to determine they are in fact bad. Also, any advice on seeing if my data is still good on the bad drives is appreciated. Thanks, Mark Before you do ANYTHING at all, you want to make absolutely certain the disks failed and that it was not something else that made it look like they failed. Do not rush to erase parity by pressing the button labeled "restore" it is a last resort, only when you have given up. First step... save a copy of the syslog. It might just have the clue to what happened. Post it on this thread. Next, power cycle. Perhaps one of the drives will come back to life. Next, open up the server and re-seat the cables. Perhaps one of the drives will come back to life. Did the two drives fail at the same time? It might be a bad power connector? or a bad "Y" splitter? or even a bad power supply where one "rail" is bad. Only when everything else fails should you press "Restore" and rebuild parity based on the other three drives. It is a last resort. Once you do that all chances of getting back anything that was on the two failed drives is forever gone. (unless one of them can come back to life) Post your syslog. (Instructions in the wiki under Troubleshooting) Do not press restore until you get feedback in this thread. By the way, it is NOT one in a million to have two drives fail. There are only two kinds of drives in the world... Those that have failed, and those that have not yet failed...but will.. just wait a bit longer. If you have been limping along with one disk failure for days/weeks/months and the other just failed... sorry... hope nothing important was on the two failed drives... only consolation, on most any other RAID array, all your data would now be lost, not just the two drives. In any case, don't rush to press the "Restore" button... you might be able to get one or both drives alive again. It might not hurt to get a replacement drive now... odds are you'll need it eventually regardless of what happens. Don't install it yet... not till you get back and post more of the details of your failure (hint... post a system log) Joe L.
June 1, 200917 yr Author I guess I should have given more info. I did check the status of the drives the day before so everything was good < 24 hours before failure. When it failed, I didn't capture the syslog so that info is gone (lesson learned). When I rebooted, the BIOS won't recognize the 2 failed drive and I can't get a separate MAC system to recognize them either. I tried a series of experiments: 1. I first plugged the bad drives into a known good SATA/power combination on this unraid box. The drives did not work. 2. I then plugged a good drive from a different sata port on the unraid to this bad port/power and it would not work. From those experiments, I concluded that both drives are dead. Any additional advice is appreciated though. Does restore rebuild the array/parity without destroying the data on the 3 good drives that are still left? Mark
June 1, 200917 yr I guess I should have given more info. I did check the status of the drives the day before so everything was good < 24 hours before failure. When it failed, I didn't capture the syslog so that info is gone (lesson learned). When I rebooted, the BIOS won't recognize the 2 failed drive and I can't get a separate MAC system to recognize them either. I tried a series of experiments: 1. I first plugged the bad drives into a known good SATA/power combination on this unraid box. The drives did not work. 2. I then plugged a good drive from a different sata port on the unraid to this bad port/power and it would not work. From those experiments, I concluded that both drives are dead. Any additional advice is appreciated though. Does restore rebuild the array/parity without destroying the data on the 3 good drives that are still left? Mark It is very unlikely for two drives to fail in a 24 hour period. (possible yes, but unlikely) What brand/model drives? There is a whole series of Seagate drives that will get their firmware confused in a way where they will not respond to a bios... There is whole thread about that... They can be fixed, and the data recovered.... with a special cable and a bit of firmware. Are they Seagate Drives? I'd say to try a new SATA cable. Make certain the other end is/are firmly plugged into the motherboard... If the BIOS does not see the drives... no way will unRAID. I'm a bit confused about this: 2. I then plugged a good drive from a different sata port on the unraid to this bad port/power and it would not work. Are you saying the good drive on the bad port would NOT work? If so, the port/and/or the cable is bad. If all else fails, if the contents can be re-ripped from media you own... it might be the only way. Pressing the button labeled "restore" will not destroy data on the working drives... It will create a new config/super.dat file based on the connected, assigned, and working drives and when the array is subsequently started, it will calculate a new set of parity on the parity disk based on those drives. Pressing it invalidates your existing parity... even if all your drives were perfect and you will not be protected by parity until a full parity calc is completed.. (Pressing restore will not do anything to restore your old data on failed disks) Joe L.
June 1, 200917 yr Author I guess I should have given more info. I did check the status of the drives the day before so everything was good < 24 hours before failure. When it failed, I didn't capture the syslog so that info is gone (lesson learned). When I rebooted, the BIOS won't recognize the 2 failed drive and I can't get a separate MAC system to recognize them either. I tried a series of experiments: 1. I first plugged the bad drives into a known good SATA/power combination on this unraid box. The drives did not work. 2. I then plugged a good drive from a different sata port on the unraid to this bad port/power and it would not work. From those experiments, I concluded that both drives are dead. Any additional advice is appreciated though. Does restore rebuild the array/parity without destroying the data on the 3 good drives that are still left? Mark It is very unlikely for two drives to fail in a 24 hour period. (possible yes, but unlikely) What brand/model drives? There is a whole series of Seagate drives that will get their firmware confused in a way where they will not respond to a bios... There is whole thread about that... They can be fixed, and the data recovered.... with a special cable and a bit of firmware. Are they Seagate Drives? The drives are the Seagate 1TB 7200.11 drives. I tried a different SATA cable with no luck either. I'd say to try a new SATA cable. Make certain the other end is/are firmly plugged into the motherboard... If the BIOS does not see the drives... no way will unRAID. I'm a bit confused about this: 2. I then plugged a good drive from a different sata port on the unraid to this bad port/power and it would not work. Are you saying the good drive on the bad port would NOT work? If so, the port/and/or the cable is bad. No, I'm saying that I connected the cables originally going to Disk2 (a good working drive) into Disk1 (failed drive) and it would not recognize it under Disk2. It just showed Disk2 as missing disk. If all else fails, if the contents can be re-ripped from media you own... it might be the only way. Pressing the button labeled "restore" will not destroy data on the working drives... It will create a new config/super.dat file based on the connected, assigned, and working drives and when the array is subsequently started, it will calculate a new set of parity on the parity disk based on those drives. Pressing it invalidates your existing parity... even if all your drives were perfect and you will not be protected by parity until a full parity calc is completed.. (Pressing restore will not do anything to restore your old data on failed disks) Joe L.
June 1, 200917 yr Author Guess I don't know how to use this quote system. The drives are Seagate 7200.11 drives - both of them. Also, to answer your second question. No, I'm saying that I connected the cables originally going to Disk2 (a good working drive) into Disk1 (failed drive) and it would not recognize it under Disk2. It just showed Disk2 as missing disk.
June 1, 200917 yr Guess I don't know how to use this quote system. The drives are Seagate 7200.11 drives - both of them. Also, to answer your second question. No, I'm saying that I connected the cables originally going to Disk2 (a good working drive) into Disk1 (failed drive) and it would not recognize it under Disk2. It just showed Disk2 as missing disk. See this thread... http://lime-technology.com/forum/index.php?topic=2773.0 and here: http://seagate.custkb.com/seagate/crm/selfservice/search.jsp?DocId=207957 and here is a huge thread describing how you can reset the drives and get them working again and NOT lose your data : http://www.msfn.org/board/index.php?s=a28fdb0dbf0bca1d27ef53a75ef62b67&showtopic=128807 Odds are good that both of your drives have locked up like this Explanation: The problem with the drive is that the drive is stuck/halt in a BUSY state. This basically locks up the SATA port, causing the BIOS problem (it cannot detect the drive at POST) I've read that Seagate will perform the data recovery service for you... (or they did when this whole mess started) Call their support line and make some inquiries if you are not up to the hardware dis-assembly/software reset repair. Joe L.
June 2, 200917 yr Author Thanks. I could easily update the firmware, but assume that's not going to work in this case since my drives are already in a confused state. I'm also not near proficient enough to reset the drive according to the thread you sent, but do appreciate your advice and assistance. I'll contact Seagate in the morning and see what they say, but I doubt they do data recovery. Thanks, Mark
June 2, 200917 yr I'll contact Seagate in the morning and see what they say, but I doubt they do data recovery. Thanks, Mark Well... they don't recover the data, but if you send them the drive, they'll reset the "busy" flag and send it back to you with your data intact. (or at least that's what I read in that thread... You'll need to inquire in the AM) If not, I'm sure the same reset service is available from second parties as well. I guess it depends on how much you lost and if it is easily replaced on those two disks of yours. Joe L.
June 2, 200917 yr Author Well... they don't recover the data, but if you send them the drive, they'll reset the "busy" flag and send it back to you with your data intact. (or at least that's what I read in that thread... You'll need to inquire in the AM) If not, I'm sure the same reset service is available from second parties as well. I guess it depends on how much you lost and if it is easily replaced on those two disks of yours. Thanks. I'll try that in the morning. Should I restore my unraid in the meantime so I can have protection and then hopefully when i get my drives back with the data on them re-connect them and re-sync parity? Does unRAID have the ability to recognize a new disk with data already on it?
June 2, 200917 yr Well... they don't recover the data, but if you send them the drive, they'll reset the "busy" flag and send it back to you with your data intact. (or at least that's what I read in that thread... You'll need to inquire in the AM) If not, I'm sure the same reset service is available from second parties as well. I guess it depends on how much you lost and if it is easily replaced on those two disks of yours. Thanks. I'll try that in the morning. Should I restore my unraid in the meantime so I can have protection and then hopefully when i get my drives back with the data on them re-connect them and re-sync parity? You can do that... Just un-assign the bad disks first, then press restore to have the array forget they ever existed. When you then start the array, parity will be completely computed as if your array was first being configured. By un-assigning them, it will be easier later to add them back in. Does unRAID have the ability to recognize a new disk with data already on it? Yes, unRAID will recognize them as having a valid reiserfs file system and will not ask you to reformat them. You will need to re-calculate parity when you do add them back. Joe L
June 2, 200917 yr Author Just wanted to give an update. I talked with Seagate this morning and they will still try to recover the data for free for these bad drives. They have a subsidiary company that will try to flash the drives and get them working - but no guarantees. Thanks for your help.
Archived
This topic is now archived and is closed to further replies.