January 20, 201115 yr Haha, well shit. My server is old, its a HP Proliant 110ML so I wouldn't be surprised if there's something funny in the bios. If I drop the drive into my desktop (current tech) and it stays at 1TB then yeah theres a problem. If I do that and it DOES show up at 2TB, what would my next steps be? My bios in the server is pretty much up to date (there's an update but it has nothing to do with this issue and I'd rather not mess with it if I don't need to) and the SATA card has no issue with it, though it happens on both the mobo and the SATA card. If I format it in windows, will PRECLEAR still function or will it assume it has been "used" and not want to function? Or can I just run it again and it wipes it all clean? (Basically does it refuse to run if it finds a NTFS partition) I really hope these aren't fake/busted, they came from Newegg and Newegg usually doesn't do that kind of nonsense. At least I kept all the boxes in case I need to mail them back I suppose. Thanks as always for the help Joe.
January 20, 201115 yr Haha, well shit. My server is old, its a HP Proliant 110ML so I wouldn't be surprised if there's something funny in the bios. If I drop the drive into my desktop (current tech) and it stays at 1TB then yeah theres a problem. That would be a quick and easy first test. Just see what its BIOS reports... no need to go any further. If I do that and it DOES show up at 2TB, what would my next steps be? My bios in the server is pretty much up to date (there's an update but it has nothing to do with this issue and I'd rather not mess with it if I don't need to) and the SATA card has no issue with it, though it happens on both the mobo and the SATA card.Replace the motherboard if the bios cannot be upgraded to one that knows about drives larger than 1TB. If I format it in windows, will PRECLEAR still function or will it assume it has been "used" and not want to function?It will not care one way or the other. Or can I just run it again and it wipes it all clean? (Basically does it refuse to run if it finds a NTFS partition)Nope, it will wipe it away. It will only refuse to clear it if it is already assigned to the array or mounted or "busy" in use. I really hope these aren't fake/busted, they came from Newegg and Newegg usually doesn't do that kind of nonsense. At least I kept all the boxes in case I need to mail them back I suppose.Yes, they are not likely to play games... but they could get a mis-labeled batch. Thanks as always for the help Joe. You are welcome. Joe L.
January 20, 201115 yr Alright, so W T F. Here's what I have done so far. Added disk to windows, reported only 1tb in windows xp (jumpered mind you) shut down, ran partition magic from DOS prompt who sees 2TB. Installed an NTFS partition across whole drive. Booted to windows. Windows sees 2TB (WHAT) Boot WD Diagnostic app, Sees 2TB. (WHATx2) Put drive back into UNRAID. Unraid sees 2TB (WHATx3) Here is an HDPARM full and what you asked for earlier (-N) along with a SMART, and what I see in PRECLEAR. /dev/sdc: max sectors = 3907027055/14715056(18446744073321613488?), HPA setting seems invalid (buggy kernel device driver?) HDPARM FULL: http://pastebin.com/TTbtRSUV SMART: http://pastebin.com/FvBkWY8s PRECLEAR: http://dl.dropbox.com/u/519591/UNRAID/Unreal.JPG So I mean, I guess I'm good if this passes preclear, right? Holy hell what a morning.
January 20, 201115 yr Alright, so W T F. Here's what I have done so far. Added disk to windows, reported only 1tb in windows xp (jumpered mind you) shut down, ran partition magic from DOS prompt who sees 2TB. Installed an NTFS partition across whole drive. Booted to windows. Windows sees 2TB (WHAT) Boot WD Diagnostic app, Sees 2TB. (WHATx2) Put drive back into UNRAID. Unraid sees 2TB (WHATx3) Here is an HDPARM full and what you asked for earlier (-N) along with a SMART, and what I see in PRECLEAR. /dev/sdc: max sectors = 3907027055/14715056(18446744073321613488?), HPA setting seems invalid (buggy kernel device driver?) HDPARM FULL: http://pastebin.com/TTbtRSUV SMART: http://pastebin.com/FvBkWY8s PRECLEAR: http://dl.dropbox.com/u/519591/UNRAID/Unreal.JPG So I mean, I guess I'm good if this passes preclear, right? Holy hell what a morning. It looks to be OK now. I would run MULTIPLE prelcears on this drive just to make doubly sure. Probably should do three at least... might not hurt to do 5...
January 20, 201115 yr I did the same to my other two drives (bought 3) and the fix also worked on them, go figure. I have 2 in the system now. I was going to do 3 for sure, do you think 5 is a little overkill? That's going to take like a week haha.
January 20, 201115 yr I do three on all my builds, and for a drive that was acting "funky" I think 5 would be a good idea.
January 20, 201115 yr Fair enough, 5 cycles underway. Unfortunately I cant knock out all at once, but at least I can get 2 done. Then I get to preclear the last one. Then I get to replace parity and rebuild. Then i get to replace a disk and rebuild. Then i get to replace a disk and rebuild. Then do a parity check. See you all in a month.
January 21, 201115 yr When I try to format a drive with the -A option, the confirmation screen still seems to show its going to do format on 63. Is this screen wrong? or it not accepting the -A properly? Pre-Clear unRAID Disk ######################################################################## Device Model: WDC WD10EARS-00MVWB0 Serial Number: WD-WCAZA0168148 Firmware Version: 50.0AB50 User Capacity: 1,000,204,886,016 bytes Disk /dev/sda: 1000.2 GB, 1000204886016 bytes 1 heads, 63 sectors/track, 31008336 cylinders, total 1953525168 sectors Units = sectors of 1 * 512 = 512 bytes Disk identifier: 0x00000000 Device Boot Start End Blocks Id System /dev/sda1 63 1953525167 976762552+ 83 Linux Partition 1 does not end on cylinder boundary. ######################################################################## Are you absolutely sure you want to clear this drive? (Answer Yes to continue. Capital 'Y', lower case 'es'):
January 21, 201115 yr When I try to format a drive with the -A option, the confirmation screen still seems to show its going to do format on 63. Is this screen wrong? or it not accepting the -A properly? Pre-Clear unRAID Disk ######################################################################## Device Model: WDC WD10EARS-00MVWB0 Serial Number: WD-WCAZA0168148 Firmware Version: 50.0AB50 User Capacity: 1,000,204,886,016 bytes Disk /dev/sda: 1000.2 GB, 1000204886016 bytes 1 heads, 63 sectors/track, 31008336 cylinders, total 1953525168 sectors Units = sectors of 1 * 512 = 512 bytes Disk identifier: 0x00000000 Device Boot Start End Blocks Id System /dev/sda1 63 1953525167 976762552+ 83 Linux Partition 1 does not end on cylinder boundary. ######################################################################## Are you absolutely sure you want to clear this drive? (Answer Yes to continue. Capital 'Y', lower case 'es'): It is telling you about the existing state of the drive. Appears to already be a valid partition at offset 63. If you allow it to move forward with the "-A" option, it will reparttion the drive to be aligned at sector 64.
January 21, 201115 yr When I try to format a drive with the -A option, the confirmation screen still seems to show its going to do format on 63. Is this screen wrong? or it not accepting the -A properly? Pre-Clear unRAID Disk ######################################################################## Device Model: WDC WD10EARS-00MVWB0 Serial Number: WD-WCAZA0168148 Firmware Version: 50.0AB50 User Capacity: 1,000,204,886,016 bytes Disk /dev/sda: 1000.2 GB, 1000204886016 bytes 1 heads, 63 sectors/track, 31008336 cylinders, total 1953525168 sectors Units = sectors of 1 * 512 = 512 bytes Disk identifier: 0x00000000 Device Boot Start End Blocks Id System /dev/sda1 63 1953525167 976762552+ 83 Linux Partition 1 does not end on cylinder boundary. ######################################################################## Are you absolutely sure you want to clear this drive? (Answer Yes to continue. Capital 'Y', lower case 'es'): It is showing you the existing drive information, including the existing partitioning, so you can make an informed decision that it is the drive you intend to clear. The existing partition starts on sector 63. It does not currently show you will now clear with the -A option, but that additional detail has been suggested for an improvement. I will be adding that a future version. It will help to eliminate questions from users being confused. (This whole advanced format stuff is confusing enough) Joe L.
January 23, 201115 yr i have a 2tb wd ears drive and i put a jumper on it. The problem im having is that every time i run preclear it gets stuck at 65% pre-read. Today i tried again and now it got stuck on 65% and unmenu went down, i left my computer and came back 5-10min later and i notice preclear was moving again and unmenu was back but in my syslog i see many errors can someone tell me if its still ok? again at 69% it stopped for 5min and unmenu was down again i have 2 drives that work perfect in the array and i have used unraid for a year now. i added my syslog and preclear smart_start some errors i see Jan 23 08:47:04 Tower kernel: ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen Jan 23 08:47:04 Tower kernel: ata2.00: failed command: READ DMA EXT Jan 23 08:47:04 Tower kernel: ata2.00: cmd 25/00:00:70:ff:da/00:01:98:00:00/e0 tag 0 dma 131072 in Jan 23 08:47:04 Tower kernel: res 40/00:00:00:4f:c2/00:00:00:00:00/00 Emask 0x4 (timeout) Jan 23 08:47:04 Tower kernel: ata2.00: status: { DRDY } Jan 23 08:47:04 Tower kernel: ata2: hard resetting link Jan 23 08:47:04 Tower kernel: ata2: softreset failed (device not ready) Jan 23 08:47:04 Tower kernel: ata2: applying SB600 PMP SRST workaround and retrying Jan 23 08:47:05 Tower kernel: ata2: SATA link up 1.5 Gbps (SStatus 113 SControl 310) Jan 23 08:47:05 Tower kernel: ata2.00: configured for UDMA/133 Jan 23 08:47:05 Tower kernel: ata2.00: device reported invalid CHS sector 0 Jan 23 08:47:05 Tower kernel: sd 1:0:0:0: [sdb] Result: hostbyte=0x00 driverbyte=0x08 Jan 23 08:47:05 Tower kernel: sd 1:0:0:0: [sdb] Sense Key : 0xb [current] [descriptor] Jan 23 08:47:05 Tower kernel: Descriptor sense data with sense descriptors (in hex): Jan 23 08:47:05 Tower kernel: 72 0b 00 00 00 00 00 0c 00 0a 80 00 00 00 00 00 Jan 23 08:47:05 Tower kernel: 00 00 00 00 Jan 23 08:47:05 Tower kernel: sd 1:0:0:0: [sdb] ASC=0x0 ASCQ=0x0 Jan 23 08:47:05 Tower kernel: sd 1:0:0:0: [sdb] CDB: cdb[0]=0x28: 28 00 98 da ff 70 00 01 00 00 Jan 23 08:47:05 Tower kernel: end_request: I/O error, dev sdb, sector 2564489072 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561134 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561135 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561136 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561137 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561138 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561139 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561140 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561141 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561142 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561143 syslog.txt smart_start_sdb.txt
January 23, 201115 yr Basically, the issue is the drive is locking up. It is also reporting an invalid geometry to the Linux kernel. Jan 23 08:47:36 Tower kernel: ata2.00: device reported invalid CHS sector 0 The kernel keeps trying to reset the disk, but that is failing. This seems to be common on the EARS drives when the jumper is added/deleted until the MBR is zeroed after a jumper change. (Or, at least that seems to fix it in most cases) Jan 23 08:47:04 Tower kernel: ata2: hard resetting link Jan 23 08:47:04 Tower kernel: ata2: softreset failed (device not ready) Jan 23 08:47:04 Tower kernel: ata2: applying SB600 PMP SRST workaround and retrying Jan 23 08:47:05 Tower kernel: ata2: SATA link up 1.5 Gbps (SStatus 113 SControl 310) Jan 23 08:47:05 Tower kernel: ata2.00: configured for UDMA/133 Jan 23 08:47:05 Tower kernel: ata2.00: device reported invalid CHS sector 0 Jan 23 08:47:05 Tower kernel: sd 1:0:0:0: [sdb] Result: hostbyte=0x00 driverbyte=0x08 Jan 23 08:47:05 Tower kernel: sd 1:0:0:0: [sdb] Sense Key : 0xb [current] [descriptor] Jan 23 08:47:05 Tower kernel: Descriptor sense data with sense descriptors (in hex): Jan 23 08:47:05 Tower kernel: 72 0b 00 00 00 00 00 0c 00 0a 80 00 00 00 00 00 Jan 23 08:47:05 Tower kernel: 00 00 00 00 Jan 23 08:47:05 Tower kernel: sd 1:0:0:0: [sdb] ASC=0x0 ASCQ=0x0 Jan 23 08:47:05 Tower kernel: sd 1:0:0:0: [sdb] CDB: cdb[0]=0x28: 28 00 98 da ff 70 00 01 00 00 Jan 23 08:47:05 Tower kernel: end_request: I/O error, dev sdb, sector 2564489072 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561134 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561135 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561136 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561137 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561138 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561139 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561140 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561141 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561142 Jan 23 08:47:05 Tower kernel: Buffer I/O error on device sdb, logical block 320561143 Jan 23 08:47:05 Tower kernel: ata2: EH complete Jan 23 08:47:35 Tower kernel: ata2.00: limiting speed to UDMA/100:PIO4 Jan 23 08:47:35 Tower kernel: ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen Jan 23 08:47:35 Tower kernel: ata2.00: failed command: READ DMA EXT Jan 23 08:47:35 Tower kernel: ata2.00: cmd 25/00:00:70:00:db/00:01:98:00:00/e0 tag 0 dma 131072 in Jan 23 08:47:35 Tower kernel: res 40/00:00:00:4f:c2/00:00:00:00:00/00 Emask 0x4 (timeout) Jan 23 08:47:35 Tower kernel: ata2.00: status: { DRDY } Jan 23 08:47:35 Tower kernel: ata2: hard resetting link Jan 23 08:47:35 Tower kernel: ata2: softreset failed (device not ready) Jan 23 08:47:35 Tower kernel: ata2: applying SB600 PMP SRST workaround and retrying Jan 23 08:47:36 Tower kernel: ata2: SATA link up 1.5 Gbps (SStatus 113 SControl 310) Jan 23 08:47:36 Tower kernel: ata2.00: configured for UDMA/100 Jan 23 08:47:36 Tower kernel: ata2.00: device reported invalid CHS sector 0
January 23, 201115 yr so i just let it preclear and it will be fine? should i do another preclear after this one? thanks
January 23, 201115 yr so i just let it preclear and it will be fine? should i do another preclear after this one? thanks Odds are it is not able to write to the drive... You'll know if it gets to the end and says it completed successfully. If there are huge amounts of errors being written to the system log and you run out of RAM because it uses it all up you'll crash your server. I'd stop the existing pre-clear, stop the array, zero the MBR, power cycle, then try once more.
January 24, 201115 yr didn't see your post, i left preclear and its on step 10 post-read. so if this read has no errors then its good?
January 24, 201115 yr didn't see your post, i left preclear and its on step 10 post-read. so if this read has no errors then its good? We'll know more when you see the final report. You can test if the pre-clear signature was successfully written to the drive with: preclear_disk.sh -t /dev/sdX
January 26, 201115 yr I am considering adding a few 2TB EARS drives it was recommend that I run Preclear first. I never ran Preclear before and wanted to see if I got everything correct after reading through this post. Last thing I want to do is screw something up. After I install 4.7 is this how I would go about replacing the Parity and Data Disk1? Not sure if it matters, but I do not have any add-ons installed, so was not not sure if any are necessary. I only ask because some people suggested unmenu so you can use the "screen" function. 1) Download preclear_disk.sh -> unzip -> add to same folder as bzroot and bzimage 2) Power down server and replace parity drive with new EARS drive 3) Power on server and open a telnet window from PC 4) Run the Preclear script "preclear_disk.sh /dev/sda" (Is "-A" no longer necessary if no jumper used on EARS in 4.7?) 5) Run at least 3 cycles 6) If Preclear is good, then assign that drive to the parity and let it do a parity check 7) Repeat above steps for for Data Disk 1 Thanks in advance.
January 26, 201115 yr I am considering adding a few 2TB EARS drives it was recommend that I run Preclear first. I never ran Preclear before and wanted to see if I got everything correct after reading through this post. Last thing I want to do is screw something up. After I install 4.7 is this how I would go about replacing the Parity and Data Disk1? Not sure if it matters, but I do not have any add-ons installed, so was not not sure if any are necessary. I only ask because some people suggested unmenu so you can use the "screen" function. 1) Download preclear_disk.sh -> unzip -> add to same folder as bzroot and bzimage Yes 2) Power down server and replace parity drive with new EARS driveNo. If you assign it as parity, the preclear script will not run on it. You must NOT assign the drive to the unRAID array. 3) Power on server and open a telnet window from PCYes 4) Run the Preclear script "preclear_disk.sh /dev/sda" (Is "-A" no longer necessary if no jumper used on EARS in 4.7?)No, if there is no jumper on the disk you do want the "-A" option. So... preclear_disk.sh -A /dev/sda 5) Run at least 3 cyclesYes. I would do one at a time, but you could type preclear_disk.sh -c 3 -A /dev/sda if you wanted to do all three at once. 6) If Preclear is good, then assign that drive to the parity and let it do a parity checkYes 7) Repeat above steps for for Data Disk 1Yes Thanks in advance. You are welcome.
January 27, 201115 yr I've been running pre clear on my new hdd and the first cycle is almost complete but it's been running for 41 hours so far and stupidly i set it to run for 3 cycles. Its now at 96% of the first cycle and honestly i dont want to run it anymore. Once this cycle is complete can i just exit pre clear?
January 27, 201115 yr I assumed as much just wanted to be sure Cheers You can test if the disk is pre-cleared by typing preclear_disk.sh -t /dev/sdX The test only takes a few seconds.
January 27, 201115 yr Just noticed the 1.1 release changes -- thanks for -W and -V A few posts back you say we no longer need to use -A if no jumper is installed -- could you please add that to release notes. Now a problem: I was in Step 2 of pre-clearing a WD Green 2 TB (previously used with content) and the server did an emergency shutdown on me. Normally I would just go ahead and restart all over, but the problem is that 'fdisk -l' is showing me strange results: Disk /dev/sde: 2000.3 GB, 2000398934016 bytes 255 heads, 63 sectors/track, 243201 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk identifier: 0x00000000 Device Boot Start End Blocks Id System Disk /dev/sdf: 2000.3 GB, 2000397852160 bytes 255 heads, 63 sectors/track, 243201 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Disk identifier: 0x00000000 Disk sde is the one I was working on, but sdf is supposed to a previously pre-cleared 1 TB Seagate HDD. I was going to add sdf to the array but can't now, even though the unRAID menu shows sdf correctly listed as expected. Should I just go ahead and redo sde and then reboot -- hopefully things will work out, or what else should I try?
Archived
This topic is now archived and is closed to further replies.