dave Posted January 15, 2015 Author Share Posted January 15, 2015 I really appreciate all your help, Gary! I have sent a PM to Tom (limetech). Hopefully he can pop in here and set me straight. Dave Link to comment
dave Posted January 15, 2015 Author Share Posted January 15, 2015 Couple ideas I had: 1) The 3TB has a "bad" version of parity on it (from when I added it too soon and triggered this whole mess). Does that matter? Does the drive need to be precleared again? - or - 2) Do I need to physically unplug the bad disk? It is currently still showing up in the drop-downs as an option to be assigned. - or - 3) Could a good old fashioned reboot have any impact? And then run through the swap-disable process again? I assume this wouldn't make anything worse! Link to comment
trurl Posted January 15, 2015 Share Posted January 15, 2015 Couple ideas I had: 1) The 3TB has a "bad" version of parity on it (from when I added it too soon and triggered this whole mess). Does that matter? Does the drive need to be precleared again? - or - 2) Do I need to physically unplug the bad disk? It is currently still showing up in the drop-downs as an option to be assigned. - or - 3) Could a good old fashioned reboot have any impact? And then run through the swap-disable process again? I assume this wouldn't make anything worse! Definitely no to 1. 2 shouldn't matter. I would wait on 3 to see if you get any other ideas. Link to comment
garycase Posted January 15, 2015 Share Posted January 15, 2015 Definitely no to #1 (as trurl noted) #2 MAY make a difference, but I'd wait for Tom's reply before doing anything else. #3 shouldn't hurt anything ... but watch carefully just in case it results in a parity check being started (and immediately abort if so). I'd probably not do anything until Tom replies to your note. Link to comment
dave Posted January 16, 2015 Author Share Posted January 16, 2015 Tried the reboot last night. The array put itself into an odd configuration: Old parity disk and disk2 unassigned. So I stopped the array, changed the parity and disk2 drives. Got the swap-disable messaging, checked the box, hit Start and the same thing happened. Is there a way to a swap-disable via command line? Do I need to upgrade my unRAID version? Is that even possible in this state? Thx, Dave Also, is there a better way to contact Lime-Tech? I've both emailed and PM'd them several days ago with no response. Link to comment
garycase Posted January 17, 2015 Share Posted January 17, 2015 Did you PM Tom directly? That's usually the best way, and unless he's usually very good at replying pretty quick. Link to comment
dave Posted January 17, 2015 Author Share Posted January 17, 2015 Did you PM Tom directly? That's usually the best way, and unless he's usually very good at replying pretty quick. Yup, I sent a PM to limetech. Three days ago now :-( Link to comment
dave Posted January 19, 2015 Author Share Posted January 19, 2015 My array has been down for over a week now. Waiting for Tom as the only option sucks. I browsed the files on the bad disk and there was nothing critical on there (I mainly care about photos as they cannot be reproduced). With that said, *IF* I swap back in the 3TB parity that had ~2,500 errors after the full parity sync and then replace the bad disk2 with another new 3TB drive - would the risk of losing data be completely limited to disk2? After that, I would try to recover the data by copying off then back on. Thoughts? Link to comment
garycase Posted January 19, 2015 Share Posted January 19, 2015 Yes, if you simply do a New Config with the new parity drive and a new 3TB drive for data and let parity build, the only risk you're taking is that you can't recover the data on disk #2. Once you've done the parity rebuild, you completely lose any possibility of rebuilding the data from the failed disk -- but anything you can actually recover from that disk is a different story. Before doing that, you may want to connect the failed disk (#2) to a SATA port on a Windows PC; load the free Linux Reader [ http://www.diskinternals.com/linux-reader/ ]; and see what data you can "see" on that disk ... copying anything you can from that disk to another storage location. That will let you know just how much you're likely to lose if you do what you suggested above. Link to comment
garycase Posted January 19, 2015 Share Posted January 19, 2015 ... For future reference, you might also want to read this thread: http://lime-technology.com/forum/index.php?topic=31020.0 Link to comment
dave Posted January 19, 2015 Author Share Posted January 19, 2015 Also, I was just browsing another thread that talked about "short-stroking" a larger drive to a smaller size. Could I do that with a 3TB drive I have to simply replace the bad 2TB disk2? Then I could do a New Config, re-assign my old 2TB parity drive, and let disk2 be rebuilt. Afterwards I could copy all the data off of the short-stroked drive to another 3TB drive (I have three not in use that are intended to be added anyways) and then "re-stroke" the one back up to 3TB. It sounds complicated but could it work? Thanks!!! Link to comment
dave Posted January 20, 2015 Author Share Posted January 20, 2015 I was reading in another read about a user swapping the parity drive and his showed up as "Wrong" too. He was instructed to stop the array, unassign the parity drive, start the array (so that the parity drive shows as missing), stop the array, then assign the parity drive. In my steps listed earlier, I only unassigned disk2 from the array and started it. Do I need to do the same for parity since it is showing as "Wrong" on the GUI? Here's the thread: http://lime-technology.com/forum/index.php?topic=31301.msg282764#msg282764 Link to comment
garycase Posted January 21, 2015 Share Posted January 21, 2015 Also, I was just browsing another thread that talked about "short-stroking" a larger drive to a smaller size. Could I do that with a 3TB drive I have to simply replace the bad 2TB disk2? Then I could do a New Config, re-assign my old 2TB parity drive, and let disk2 be rebuilt. Afterwards I could copy all the data off of the short-stroked drive to another 3TB drive (I have three not in use that are intended to be added anyways) and then "re-stroke" the one back up to 3TB. It sounds complicated but could it work? Thanks!!! You can use the SetMax feature of HDAT2 to set the drive to the correct size for a 2TB drive ... and then it'll work as a 2TB drive in UnRAID (or anything else). But to reset it to the max size, you'll lose everything that's on it -- so be aware of that. http://www.hdat2.com/ Link to comment
dave Posted January 22, 2015 Author Share Posted January 22, 2015 Ok, I can't wait any longer for Lime Tech to help. I want to get my array back up and useable. I have the following disks: 1) Original Parity (2TB) 2) Disk1 with data on it that I want to keep (2TB) 3) Disk2 has been removed from the server 4) Disk3 with data on it that I want to keep (1.5TB) 5) Disk4 with data on it that I want to keep (1TB) 6) New drive that has been precleared (3TB) 7) New drive that is currently preclearing (3TB) What are my options to get the array running as-is? I'm okay with having disk2 out of the array - I'll recover that data later. Can I do the following? 1) Stop array (it's already stopped) 2) Do New Config 3) Assign precleared 3TB to parity 4) Assign disks 1, 3, and 4 to spots 1, 2, 3 5) Start the array 6) Let unRAID build new parity 7) Should I run a parity check afterwards? Then, once the other 3TB drive is done preclearing I can add it to the array in spot 4. After it's assigned I can copy the data from the old disk2 to the array. Does anyone see any problems with that? Thanks! Dave Link to comment
trurl Posted January 22, 2015 Share Posted January 22, 2015 Ok, I can't wait any longer for Lime Tech to help. I want to get my array back up and useable. I have the following disks: 1) Original Parity (2TB) 2) Disk1 with data on it that I want to keep (2TB) 3) Disk2 has been removed from the server 4) Disk3 with data on it that I want to keep (1.5TB) 5) Disk4 with data on it that I want to keep (1TB) 6) New drive that has been precleared (3TB) 7) New drive that is currently preclearing (3TB) What are my options to get the array running as-is? I'm okay with having disk2 out of the array - I'll recover that data later. Can I do the following? 1) Stop array (it's already stopped) 2) Do New Config 3) Assign precleared 3TB to parity 4) Assign disks 1, 3, and 4 to spots 1, 2, 3 5) Start the array 6) Let unRAID build new parity 7) Should I run a parity check afterwards? Then, once the other 3TB drive is done preclearing I can add it to the array in spot 4. After it's assigned I can copy the data from the old disk2 to the array. Does anyone see any problems with that? Thanks! Dave That sounds OK. One thing you might pay attention to since you are changing slots is whether or not any or your shares are set to include or exclude drives. If so then you will need to make the appropriate changes. Definitely run a parity check after the parity build. In fact, unRAID will say that parity hasn't been checked, and until you do you won't know for sure that it was built OK. Have you tried to read the failed drive on your PC yet? Link to comment
dave Posted January 22, 2015 Author Share Posted January 22, 2015 That sounds OK. One thing you might pay attention to since you are changing slots is whether or not any or your shares are set to include or exclude drives. If so then you will need to make the appropriate changes. Definitely run a parity check after the parity build. In fact, unRAID will say that parity hasn't been checked, and until you do you won't know for sure that it was built OK. Have you tried to read the failed drive on your PC yet? Ok, thanks for confirming. I have ordered an external hard drive enclosure that should arrive tomorrow to try and read the failed disk. *fingers crossed* Link to comment
dave Posted January 27, 2015 Author Share Posted January 27, 2015 In case anyone is still following this thread, here's an update: 1) I have the array back up and running after doing a New Config and adding in my two new 3TB drives. 2) I bought an external drive dock, plugged it into my Windows 7 VM, and read it using Linux Reader. It is currently copying over the 1.7TB of data from my failing drive. Only 1.4TB to go... Does anyone know how the errors will be handled during file copy? Will the file be copied over incorrectly or will the copy fail altogether? Or is that not possible to tell?? Link to comment
itimpi Posted January 27, 2015 Share Posted January 27, 2015 Does anyone know how the errors will be handled during file copy? Will the file be copied over incorrectly or will the copy fail altogether? Or is that not possible to tell?? I think the most likely thing on error is that a particular file will generate an error if it cannot be read. However there is always a faint chance that the drive will report data has been read correctly when it has not. This latter case is the classic one (albeit rare) that can lead to 'bit-rot' of data, and if you want to protect against this then you need to have file checksums for each file so you can detect it. Link to comment
garycase Posted January 27, 2015 Share Posted January 27, 2015 Does anyone know how the errors will be handled during file copy? Will the file be copied over incorrectly or will the copy fail altogether? Or is that not possible to tell?? Agree with itimpi -- it's not really possible to tell if you don't have file checksums. But by far the most likely thing is you'd get a read error reported and the offending file wouldn't copy. Note that if you're using Windows Explorer to do the copies, this will also abort the entire copy process (not good). If that happens, use something like SyncBack (or TeraCopy) to start the copy over ... these will continue copying the good files if a single file has an error. Link to comment
dave Posted January 27, 2015 Author Share Posted January 27, 2015 Does anyone know how the errors will be handled during file copy? Will the file be copied over incorrectly or will the copy fail altogether? Or is that not possible to tell?? Agree with itimpi -- it's not really possible to tell if you don't have file checksums. But by far the most likely thing is you'd get a read error reported and the offending file wouldn't copy. Note that if you're using Windows Explorer to do the copies, this will also abort the entire copy process (not good). If that happens, use something like SyncBack (or TeraCopy) to start the copy over ... these will continue copying the good files if a single file has an error. I am using the Export Wizard in the Linux Reader program that you mentioned. That should be fine right? See attached. Link to comment
garycase Posted January 27, 2015 Share Posted January 27, 2015 Yes, that should be fine -- and I'm fairly certain it will alert you if there are any errors encountered. Link to comment
dave Posted February 4, 2015 Author Share Posted February 4, 2015 Final update: I got everything copied off of the bad drive. Only one file gave me an error, so everything worked out well. My array is back up and running flawlessly. Thanks everyone for your help! Link to comment
garycase Posted February 4, 2015 Share Posted February 4, 2015 Glad it worked out well with almost no data loss. Now it's time to fix this: ... I screwed up and did not do a backup... ... a few thoughts about that: http://lime-technology.com/forum/index.php?topic=31020.0 Link to comment
dave Posted February 4, 2015 Author Share Posted February 4, 2015 Glad it worked out well with almost no data loss. Now it's time to fix this: ... I screwed up and did not do a backup... ... a few thoughts about that: http://lime-technology.com/forum/index.php?topic=31020.0 Oh ya, I've already got a backup of my flash drive! Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.