October 26, 200916 yr I don't know if anyone can help. As everyone knows for the last month or so I have had massive hard drive problems one after another. The last problem was drive corruption on drive 2. That was corrected & I pulled the drive out & replaced it with another 500 GB drive. I told unRAID to restore because I didn't want to rebuild the drive from the failed drive. There wasn't any data present. Because I had copied everything off of this drive to drive 15. During the parity sync which followed the restore operation, disaster struck again. Drive 15 which was just replaced from Seagate within the last 2 weeks smoked. The circuit board fried. The drive is gone. Now unRAID says parity is invalid & drive 15 is missing. Unfortunatly a syslog was not created. I had the tail command running & nothing was generated. When the drive fried the power supply shut down. I would guess since parity sync was in progress that the old parity data is gone. I hope I am wrong but this is what I suspect. Any help would be appreciated.
October 26, 200916 yr I don't know if anyone can help. As everyone knows for the last month or so I have had massive hard drive problems one after another. The last problem was drive corruption on drive 2. That was corrected & I pulled the drive out & replaced it with another 500 GB drive. I told unRAID to restore because I didn't want to rebuild the drive from the failed drive. There wasn't any data present. Because I had copied everything off of this drive to drive 15. During the parity sync which followed the restore operation, disaster struck again. Drive 15 which was just replaced from Seagate within the last 2 weeks smoked. The circuit board fried. The drive is gone. Now unRAID says parity is invalid & drive 15 is missing. Unfortunatly a syslog was not created. I had the tail command running & nothing was generated. When the drive fried the power supply shut down. I would guess since parity sync was in progress that the old parity data is gone. I hope I am wrong but this is what I suspect. Any help would be appreciated. Did you press the button labeled "Restore" ? It sounds like you did. It basically re-names your config/super.dat file to confg/super.old. That file holds the superblock that tracks the installed drives in your array and the array status. The actual "parity" drive marked as not-valid, and the process of re-calculating parity with the new disk configuration (those currently working and assigned) is started. Once it starts, you do not have parity protection until it finishes. You have part of the parity drive calculated based on the old disk2, and part with the new disk2. You can try to force it to think only disk15 is bad with a modified "trust my disk" procedure setting the invalidslot to 15. Then when you start the array it might let you see what was on the disk... it all depends on what was on dsk2. Basically you had a two disk failure... just be happy it was not raid5, or you would now have lost all your data. Joe L.
October 26, 200916 yr Author New update - Just got off of the phone with Seagate. They will not replace this drive at this time. They claim the warranty is out. This is a 1.5Tb internal retail drive with a 5 year warranty. Of course I have the replacement drive - so it is a recertified drive. They claim they have to try & get a new PN for it & maybe by the end of the week they may be able to replace the drive. B.S. I will never buy another Seagate again...
October 26, 200916 yr New update - Just got off of the phone with Seagate. They will not replace this drive at this time. They claim the warranty is out. This is a 1.5Tb internal retail drive with a 5 year warranty. Of course I have the replacement drive - so it is a recertified drive. They claim they have to try & get a new PN for it & maybe by the end of the week they may be able to replace the drive. B.S. I will never buy another Seagate again... 1.5TB drives have not been out for 5 years... I don't care if they replaced it once already, that should not reduce the original term of 5 years. I'm confused... and you should be livid, and back on the phone with them... Joe L.
October 26, 200916 yr What I need to try to do is to make parity valid.. As I said, you can try a modified version of the "trust my parity" procedure as described in the wiki, but instead of setting slot 99 in the array as invalid, set slot 15 as invalid. You really have nothing to lose... give it a try. Joe L.
October 26, 200916 yr Author New update - Just got off of the phone with Seagate. They will not replace this drive at this time. They claim the warranty is out. This is a 1.5Tb internal retail drive with a 5 year warranty. Of course I have the replacement drive - so it is a recertified drive. They claim they have to try & get a new PN for it & maybe by the end of the week they may be able to replace the drive. B.S. I will never buy another Seagate again... 1.5TB drives have not been out for 5 years... I don't care if they replaced it once already, that should not reduce the original term of 5 years. I'm confused... and you should be livid, and back on the phone with them... Joe L. I'm more than livid.. And My wife is gonna kill me if I can't recover drive 15... Seagate kept telling me that the original drive was mislabled as a internal drive.. I told her that it was not mislabeled because I bought it at Best Buy as a Retail Box Internal Drive. They said they have to research that & in a few days they would have to give it a new PN number.. They relalize that it was there fault but that doesn't make it any better..
October 26, 200916 yr Author When I do the "Trust My Parity" procedure - I should follow everything except hitting the restore button... Right?
October 26, 200916 yr When I do the "Trust My Parity" procedure - I should follow everything except hitting the restore button... Right? Wrong. It is one of the very few times when you do want to press the button labeled "Restore", but DO NOT start the array until AFTER you type the mdcmd set invalidslot 15 command. Once you see the expected response from it you can then press the "Start" button. Joe L.
October 26, 200916 yr Author I guess this will not work.. Parity must not be right. All drives are green, except 15 is not there.
October 26, 200916 yr Here is the syslog It looks as if you have not got drive15 assigned to the array. Stop the array... Do not write to it for the moment, look at the device assignments page. What does it show for disk15?
October 26, 200916 yr Author There is no drive assigned. I had to pull drive 15 out because it was smoking. I made a copy of the super.dat before I did the "trust my parity" procedure. What I need to do is have parity valid & then have it imitate drive 15. Recreate drive 15 on the fly. I know this is probably impossible. pci-0000:001f.2-scsi-0:0:0:0 (no device)
October 26, 200916 yr There is no drive assigned. I had to pull drive 15 out because it was smoking. I made a copy of the super.dat before I did the "trust my parity" procedure. What I need to do is have parity valid & then have it imitate drive 15. Recreate drive 15 on the fly. I know this is probably impossible. pci-0000:001f.2-scsi-0:0:0:0 (no device) Smoking is not good... (too young, bad for health, etc) Did you save a copy of the disk.cfg file? You need to convince it that drive15 should exist, even though it is missing. You would need it You can certainly try copying back the old super.dat back. Look in config/disk.cfg
October 26, 200916 yr Author OK I'm back to invalid parity & disk 15 missing like in the first screenshot. I reverted back to my old super.dat file. I've attached the new syslog & a copy of my disk.cfg file.
October 27, 200916 yr Author Now it's just getting better. Looks like the power supply is now failing. Guess the drive smoking took out the supply. The machine will not boot unless I unplug a few drives. I've got a RMA started with Corsair. It is a VX550W supply. Hopefully they will replace it. I bought the supply on 1-28-2008. Thanks for the help... P.S. In the mean time if anyone has any ideas on how to get drive 15 back let me know... Phil
October 29, 200916 yr Guess the drive smoking took out the supply. It may be the other way around, the failing PSU may have been what was causing drives to fail. I really don't know which one it is, but the implications are important. If it is a failing PSU that is killing drives, then you will want to use your array as little as possible, to avoid 'cooking' another drive. In the mean time if anyone has any ideas on how to get drive 15 back let me know... There are 2 ways that I can think of. The easiest would be that you still have a disk.cfg and super.dat saved when all of the current drives were installed AND were green. Then you could restore these 2 files and your array would be usable, with Disk 15 a virtual drive. The second requires another drive, installed as Disk 15, to which you would rebuild with the "invalid slot 15" method that Joe mentioned. It failed before, because neither this nor the Trust My Array method can work if a drive is missing. The step to press the Restore button causes unRAID to forget any past drives, and setup a new array with only the drives that it can *currently* see. I'm really sorry for all the trouble you have been having, and hope there is no further data loss. I don't know for sure what your next steps should be. It does seem imperative that you install a trustworthy power supply, before trying *anything* else. And a saved config from before this trouble could help, to at least get the most vital data saved elsewhere.
October 29, 200916 yr Author I've unplugged this box. I've sent the power supply back to corsair for replacement. They approved my RMA request. You're probably right about the power supply taking out the hard drives. Everything I lost was only movie rips. I can rerip everything. I had a couple of home movies on that drive but I also have them on DVD so no problem. I have since reused my 2Tb parity drive in a different machine, so parity is lost. I'm using it in my SageTV HTPC as a parity drive for my TV recordings. I am using FlexRaid for that. My main focus now is not loosing any TV recordings. The wife & kid would have my hide. I do thank everyone for their help & concern.
October 29, 200916 yr Author New update - Just got off of the phone with Seagate. They will not replace this drive at this time. They claim the warranty is out. This is a 1.5Tb internal retail drive with a 5 year warranty. Of course I have the replacement drive - so it is a recertified drive. They claim they have to try & get a new PN for it & maybe by the end of the week they may be able to replace the drive. B.S. I will never buy another Seagate again... 1.5TB drives have not been out for 5 years... I don't care if they replaced it once already, that should not reduce the original term of 5 years. I'm confused... and you should be livid, and back on the phone with them... Joe L. I'm more than livid.. And My wife is gonna kill me if I can't recover drive 15... Seagate kept telling me that the original drive was mislabled as a internal drive.. I told her that it was not mislabeled because I bought it at Best Buy as a Retail Box Internal Drive. They said they have to research that & in a few days they would have to give it a new PN number.. They relalize that it was there fault but that doesn't make it any better.. Got off of the phone with Seagate. They have now fixed the problem with the drives label problem. They are now shipping another 1.5Tb drive to me today. Should get it in a day or two. Looks like we are finally getting all of the hardware replaced under warranty.
October 31, 200916 yr Author The new Seagate 1.5Tb drive arrived today. I live basically in the same town they have a RMA center, so I get them quick. Now waiting for the power supply..
Archived
This topic is now archived and is closed to further replies.