January 23, 201313 yr Today I ripped a Blu-ray - "The Illusionist". We had/have the DVD and liked it so decided to upgrade to Blu-ray after finding a cheap disc at Half Price Books. I normally rip the Blu-rays into an MKV container. I use a 2 step process with Eac3to and mkvmerge. This method has always worked well for me. I usually set the destination for the MKV file to a folder on my unRAID server. However, today I ran into some problems. While mkvmerge was running, it stopped and reported that it could no longer access the destination folder. So I changed the destination to a local folder. It worked fine writing to my local drive. Then I just tried moving the MKV (17.5 GB) to my unRAID server. I have 4 user shares: 1. Movies Allocation method: high-water Min. free space: 0 Split level: 1 Included disk(s): blank Excluded disk(s): disk1 2. Music Allocation method: high-water Min. free space: 0 Split level: 3 Included disk(s): blank Excluded disk(s): disk2 3. Photos Allocation method: high-water Min. free space: 0 Split level: 1 Included disk(s): disk4 Excluded disk(s): disk1, disk2 4. Videos Allocation method: high-water Min. free space: 0 Split level: 3 Included disk(s): blank Excluded disk(s): disk1 In addition to my parity drive, I have 6 drives - disk1 - disk6. All are 2TB drives, except drives disk5 and disk6 are both 1TB drives. Disk1 has 584 GB free. Disk2 has 32 GB free. Disk3 has 73 GB free. Disk4 has 490 GB free. Disk5 has 123 GB free. Disk6 has 976 GB free. As you can see there's plenty of room for this 17.5 GB file. Moving the file to a folder in the Movies share didn't work. Using Windows Explorer, I simply cut and paste to the server. The transfer speed was painfully slow - 1.3MB/s. After, waiting several hours, the transfer just crapped out and finally stopped before finishing. I got a message saying drive no longer found or something like that. A few weeks ago I had to replace my router. I never bothered to restart the server, because the new router didn't have any trouble finding it. Besides, it's been working fine since and I haven't noticed any problems writing to the Movies share or slow transfer speeds. So, after generating a syslog, I restarted unRAID. I tried moving the file again. Again, it was really slow, but this time, the transfer stopped only after a few minutes. Next, I tried moving the file to the Videos share. That worked fine. Transfer speed was over 32MB/s. After moving the file to my Videos share. I am now moving it to the Movies share. So far, it is working (only been a few minutes) and the transfer speed is better, 14MB/s, but still seems pretty slow since the transfer isn't over the network. Is there anything wrong with the way my shares are configured? Sent from my Nexus 7 using Tapatalk 2 syslog-2013-02-23.txt
January 24, 201313 yr Author On my server disk4 is sdg and I've attached the smart report to this post. Thanks for taking the time to help me out. It would be a bummer if this drive is failing. It's my newest drive. smart-disk4_sga_2013-02-23.txt
January 24, 201313 yr The drive has unreadable sectors. I needs to be rebuilt. If you have a spare use it now. To rebuild on the current drive use this procedure: 1. Stop the array. 2. Unassign the disk. 3. Optionally pre-clear the disk. 4. Start the array. 5. Stop the array and re-assign the disk. 6. Start the array and the disk will rebuild. You may need to also click a button to start the rebuild. Current_Pending_Sector 0x0022 100 100 000 Old_age Always - 9 The Current_Pending_Sector RAW_VALUE must be zero.
January 24, 201313 yr Author The disk was precleared before it was added to the array a few months ago. Is there any advantage to doing it again? Sent from my Nexus 7 using Tapatalk 2
January 24, 201313 yr The disk was precleared before it was added to the array a few months ago. Is there any advantage to doing it again? Sent from my Nexus 7 using Tapatalk 2 yes, you want to write to those 9 un-readable sectors that are pending re-allocation and identify any others.
January 24, 201313 yr The rebuild will write to all of the sectors causing the 9 pending to be reallocated, but new ones may appear. Pre-clear will write to the sectors and then read all of the sectors to identify any new ones.
January 24, 201313 yr Author I don't have a spare drive, so I'll run the preclear and rebuild the array. I may have to wait until after the weekend, because my kids were planning to watch movies with their friends. Is there a way to determine if I will need to replace it in the near future? In other words, does this look like the early stages of a failing drive? I'm just wondering if I should buy a spare/replacement and start the RMA process.
January 24, 201313 yr The drive may be fine after the sectors are reallocated. You should have a spare in any case.
January 27, 201313 yr Author I'm currently preclearing disk4. While I'm waiting, I was wondering why I should start the array in the 4th step. The array is stopped and disk4 is unassigned (and preclearing). Wouldn't it make sense (after finishing the preclear) to just re-assign disk4 and start the array? If I start the array in step 4, won't it automatically try rebuilding the array and remove the data that was on disk4?
January 28, 201313 yr I'm currently preclearing disk4. While I'm waiting, I was wondering why I should start the array in the 4th step. The array is stopped and disk4 is unassigned (and preclearing). Wouldn't it make sense (after finishing the preclear) to just re-assign disk4 and start the array? If I start the array in step 4, won't it automatically try rebuilding the array and remove the data that was on disk4? you must start the array with the disk un-assigned to get unRAID to forget the model/serial number of the disk so that it will accept it as its own replacement once you re-assign it..
January 29, 201313 yr Author It finished rebuilding my array some time last night. Is there a way to determine if the previously unassigned sectors are now assigned on disk4? What should I look for in my SMART report to determine if it is ok? smart01-29-13.txt
January 29, 201313 yr It finished rebuilding my array some time last night. Is there a way to determine if the previously unassigned sectors are now assigned on disk4? What should I look for in my SMART report to determine if it is ok? The drive looks OK now. It recently did find a number of corrupted sectors, but none of them were physically damaged, so all were rewritten correctly. Current_Pending_Sector is now zero, Reallocated_Sector_Ct is still zero. If you are still in doubt, you can Preclear one more time, just to be sure.
February 5, 201313 yr Author It looks like this problem has raised its ugly head again. Trying to write to the server has again failed. Mkvmerge, which usually only takes about 30 minutes for a Blu-ray, was taking over 2 hours when it finally responded that it could no longer find the destination drive. I then ran a SMART report on all of my drives. Only 2 drives reported any issues with pending sectors: Disk2 (sde): 197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 74 198 Offline_Uncorrectable 0x0030 200 200 000 Old_age Offline - 1 199 UDMA_CRC_Error_Count 0x0032 200 200 000 Old_age Always - 0 200 Multi_Zone_Error_Rate 0x0008 200 200 000 Old_age Offline - 11 Disk4 (sdg): 197 Current_Pending_Sector 0x0022 100 100 000 Old_age Always - 1 198 Offline_Uncorrectable 0x0008 100 100 000 Old_age Offline - 0 199 UDMA_CRC_Error_Count 0x000a 200 200 000 Old_age Always - 0 So, should I follow the same procedure for disk2, that I previously followed for disk4? What does "Multi Zone Error Rate" mean? And now disk4 is showing 1 pending sector. After preclearing and rebuilding about a week ago, there were 0 pending sectors. Is there any reason to be concerned with this pending sector reappearing? syslog-2013-02-05.txt smart_disk2-2013-02-05.txt smart_disk4-2013-02-05.txt
February 6, 201313 yr Author Does anyone know if a drive with 74 current pending sectors is a failing drive? When this happened to disk4 in my array, there were only 9 pending sectors. In comparison, 74 seems like a lot. Or is this still a normal amount? Should I contact Western Digital and ask for an RMA? This drive was just purchased last October. The Hitachi drive, disk4, is also still under warranty. When I create a SMART report for either of these drives the overall-health self-assessment test result is "healthy". What about only 1 pending sector, which is what I have with disk4. Will I need to repeat the stop/unassign/preclear/start/stop/start/rebuild cycle again for disk4?
February 6, 201313 yr From what I understand, disks have many spare sectors for re-allocation (many more than your showing). The way to determine would be if the raw pending or reallocated counts continue to increase. If the self-assessment result is healthy, then personally, I'd try that. Alternatively, if the RMA process for the store is straighforward enough and the disk is not needed, then you may want to try that. (I'm not sure what WD would look for to agree a disk is dead, eg - they may not accept it as failing if the count is low and the drive reporting as healthy.) EDIT: Re-reading some of the older posts, as this is a recurring issue, it may be a little more concerning and I personally, would swap the drive out (if possible) and do some further tests on it.
February 6, 201313 yr Author I don't understand how the drives are healthy when you can't write to them. Or, the transfer speed is so slow. It doesn't seem to be a good solution if you have to perform maintenance that takes about 48 hours before you have normal performance again and then you have to repeat the process every few weeks. The drive with the reoccurring problem is disk4. When I first discovered the performance problems, disk4 was reporting 9 pending sectors. I then took disk4 offline, precleared it, and then brought back online. After that process, the SMART report reported 0 (no) pending sectors. Now, a few weeks later, disk4 has 1 pending sector. Again, I'm not sure if this is the sign of a drive with a chronic problem. However, disk2 with 74 pending sectors seems to be as big or bigger problem. I just don't know if the manufacturers will agree. If my solution is to go through the offline/preclear/rebuild... cycle every few weeks with these drives. The server is going to be offline quite a lot, considering there are 2 disks and you can't preclear them both at the same time. Buying a new drive seems like an expensive solution when both of these drives are still under warranty.
February 6, 201313 yr Author I also noticed that the reallocated sector counts for my drives are: Disk0 - 0 Disk1 - 4 Disk2 - 0 Disk3 - 0 Disk4 - 0 Disk5 - 0 Disk6 - 0 Is there a different problem with disk1 that I should be aware of?
February 6, 201313 yr I also noticed that the reallocated sector counts for my drives are: Disk0 - 0 Disk1 - 4 Disk2 - 0 Disk3 - 0 Disk4 - 0 Disk5 - 0 Disk6 - 0 Is there a different problem with disk1 that I should be aware of? No.
February 6, 201313 yr It looks like this problem has raised its ugly head again. Trying to write to the server has again failed. Mkvmerge, which usually only takes about 30 minutes for a Blu-ray, was taking over 2 hours when it finally responded that it could no longer find the destination drive. I then ran a SMART report on all of my drives. Only 2 drives reported any issues with pending sectors: Disk2 (sde): 197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 74 198 Offline_Uncorrectable 0x0030 200 200 000 Old_age Offline - 1 199 UDMA_CRC_Error_Count 0x0032 200 200 000 Old_age Always - 0 200 Multi_Zone_Error_Rate 0x0008 200 200 000 Old_age Offline - 11 Disk4 (sdg): 197 Current_Pending_Sector 0x0022 100 100 000 Old_age Always - 1 198 Offline_Uncorrectable 0x0008 100 100 000 Old_age Offline - 0 199 UDMA_CRC_Error_Count 0x000a 200 200 000 Old_age Always - 0 So, should I follow the same procedure for disk2, that I previously followed for disk4? What does "Multi Zone Error Rate" mean? And now disk4 is showing 1 pending sector. After preclearing and rebuilding about a week ago, there were 0 pending sectors. Is there any reason to be concerned with this pending sector reappearing? Both these disks need to be rebuilt. The problem is that the unreadable sectors on one drive prevents the reconstruction of the other. Backup the data on each of those drives to good drives. The drives are probably fine and this issue is cause by some other component. Run memtest overnight. What model PSU is in the system?
February 6, 201313 yr Author I've ordered a new 2TB WD WD20EURS drive. I'll add it to my array and try to move everything from these drives over to other drives. Not sure how else to actually back these up. In the mean time, I'm preclearing disk2. Won't be able to run memtest until that is finished. I haven't run memtest since I first built this server, which was in the fall of 2008. If I recall correctly, you can't run memtest after you've booted to unRAID. This will be a little bit of a challenge because I no longer have a monitor that will connect to the server. I'll need to go out and buy a cable adapter. The adapters that ship with graphics cards all have the wrong gender. My hardware includes: Antec Nine Hundred mid-tower case Thermaltake Purepower W0100RU 500W ATX PSU ABIT AB9 Pro LGA 775 motherboard Intel Pentium E2160 Allendale 1.8 GHz LGA 775 CPU 2 GB Crucial Ballistix 240-pin DDR2 SDRAM (this might've been swapped out with something else. I seem to remember getting a bad stick and going through an RMA. Might of ended up with some other RAM) What makes you think this problem is caused by other components?
February 6, 201313 yr Author Oh noooo!! Did I just screw up by unmounting and preclearing one of these drives??? If the unreadable sectors on both drives prevents the reconstruction of the other, then I can't reconstruct either of these drives until they are both precleared...? So I would need to completely back up both these drives first. Then once preclearing is complete and the drives are remounted and added to the array, I could safely restore the data back to those drives. Having failed that, I may have lost all the data on both disk2 and disk4. Is that right? And you can only reconstruct after a single drive has failed, not 2. Am I screwed???
February 6, 201313 yr A single drive can be rebuilt if all of the other disks are readable. The rebuild of disk 4 might work. It could take a long time. If it doesn't work, use dd to duplicate the entire drive partition. The new copy can then be used to rebuild the missing one.
February 7, 201313 yr Author I'm not sure I understand what to do. Currently, my server is stopped. I had 2 drives with pending sectors - disk2 with 74 and disk4 with 1. I stopped the array. Then I unmounted disk2 and starting preclearing disk2. I haven't done anything to disk4, yet. So, based on your response, I might be able to complete the preclearing of disk2, then start the array, stop the array and re-assign disk2. Then when I start the array, disk2 should rebuild, but it might take a long time. Or, was this only possible with disk4? I'm not familiar with 'dd' for duplicating drive partitions (not a linux guy). Could you point me to more details? I searched the wiki and couldn't find anything specific about dd. Thanks for your help.
February 7, 201313 yr Author dgaschk- You said a single drive can be rebuilt if all other disks are readable. Is a disk readable if it has pending sectors? Then you go onto say, "the rebuild of disk4 might work." Unfortunately, that advice was after I already started preclearing disk2. Or, are you saying that it still might be possible to rebuild disk4 even though I started preclearing disk2? I'm still on step 2 of 10 in the preclearing process (copying zeros to remainder of disk to clear it). So I'll let it finish up. Can someone please tell me how to recover from this sort of disaster? Is there a way to list files on my server by disk? I know whats located in any given share, but how do I find out what folders and files are stored on each disk? Does anyone else know anything about dd? And could tell me how to use it?
Archived
This topic is now archived and is closed to further replies.