goinsnoopin Posted February 5, 2011 Share Posted February 5, 2011 I just successfully replaced my parity drive, rebuilt parity and performed a parity check with no errors. How long should I wait before I preclear the old parity drive and use it to replace one of my other data drives? Or is there anything else I should do or test before preclearing the old parity drive? Thanks, Dan Quote Link to comment
garycase Posted February 6, 2011 Share Posted February 6, 2011 You've already replaced the drive and confirmed that parity is good, so there's no reason to keep the old parity drive. Go ahead and add it to the array ... either as a new drive or to replace an existing drive. Quote Link to comment
SSD Posted February 6, 2011 Share Posted February 6, 2011 No need to preclear your old parity drive if you are going to use it to replace (upsize) another disk in the array. Quote Link to comment
T3nchu Posted February 6, 2011 Share Posted February 6, 2011 Yes, you can do it without losing data. Try this: hdparm -N p1953525168 /dev/sdf Follow it with hdparm -N /dev/sdf to see if it took effect. I had ran into HPA problem which did not surfaced until I updated to 4.7. I have a friend also had the same problem. I had 2 drives that reported problem after I updated to 4.7. I follow the above instructions and was able to correct one of the drive but the other drive is WD EARS 1.5TB without jumper on it (did not know about the jumper until now). After I issue the commends to the WD drive it came back with only 409GB on the drive and I can't seems to be able to fix it. /dev/sde: setting max visible sectors to 2930277168 (permanent) max sectors = 799570736/11041584, HPA setting seems invalid root@Tower:~# hdparm -N /dev/sde /dev/sde: max sectors = 799570736/11041584, HPA setting seems invalid root@Tower:~# hdparm -i /dev/sde /dev/sde: Model=WDC WD15EADS-00P8B0 , FwRev=01.00A01, SerialNo= WD-WCAVU0423882 Config={ HardSect NotMFM HdSw>15uSec SpinMotCtl Fixed DTR>5Mbs FmtGapReq } RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=50 BuffType=unknown, BuffSize=32767kB, MaxMultSect=16, MultSect=?16? CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=799570736 IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120} PIO modes: pio0 pio3 pio4 DMA modes: mdma0 mdma1 mdma2 UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5 udma6 AdvancedPM=no WriteCache=enabled Drive conforms to: Unspecified: ATA/ATAPI-1,2,3,4,5,6,7 root@Tower:~# hdparm -I /dev/sde /dev/sde: ATA device, with non-removable media Model Number: WDC WD15EADS-00P8B0 Serial Number: WD-WCAVU0423882 Firmware Revision: 01.00A01 Transport: Serial, SATA 1.0a, SATA II Extensions, SATA Rev 2.5 Standards: Supported: 8 7 6 5 Likely used: 8 Configuration: Logical max current cylinders 16383 16383 heads 16 16 sectors/track 63 63 -- CHS current addressable sectors: 16514064 LBA user addressable sectors: 268435455 LBA48 user addressable sectors: 799570736 device size with M = 1024*1024: 390415 MBytes device size with M = 1000*1000: 409380 MBytes (409 GB) I accidentally reformat the drive that I correct so I am in the process of rebuilding it. I had to unassign the WD EARS and initconfig the server to force it to resync. My question is, is there a way to fix my WD EARS drive and rebuild the data somehow? I also order a different MB which will prevent HPA from being a problem in the future. Update: Further observation showed there are now 2 partitions on the WD EARS drive now?? I was able to mount one of the partition and access some of the data but I get errors Feb 7 12:31:42 Tower kernel: attempt to access beyond end of device Feb 7 12:31:42 Tower kernel: sde1: rw=0, want=1423801160, limit=799570673 Feb 7 12:31:42 Tower kernel: REISERFS error (device sde1): vs-13070 reiserfs_read_locked_inode: i/o failure occurred trying to find stat data of [5 1323 0x0 SD] root@Tower:~# fdisk -l /dev/sde Disk /dev/sde: 409.3 GB, 409380216832 bytes 1 heads, 63 sectors/track, 12691598 cylinders Units = cylinders of 63 * 512 = 32256 bytes Disk identifier: 0x00000000 Device Boot Start End Blocks Id System /dev/sde1 2 46512303 1465137496 83 Linux Partition 1 does not end on cylinder boundary. So the question now is how did I get the 2 partitons with the original commend and how do i make it into 1 partition again? Thanks for the reply! Quote Link to comment
Hex Posted February 8, 2011 Share Posted February 8, 2011 Same ol story. Updated to 4.7, samsung drive reports too small. Went back to 4.6 and am going to wait until the expensive program I bought can fix the problem. I am tired of patching program errors. Did I buy into a hobby program or a professional program? I am glad the program works good as is and I guess I will not update. Post a fix subject title in the forum that THE program can fix this problem and I will try again. Thanks Quote Link to comment
PeterB Posted February 8, 2011 Share Posted February 8, 2011 Post a fix subject title in the forum that THE program can fix this problem and I will try again. Have you actually read up on the background, and understood the root of this problem? Your motherboard makes a hidden reservation of a chunk of the hard drive and you expect the application to retrieve the space??? Quote Link to comment
BRiT Posted February 8, 2011 Share Posted February 8, 2011 Hex, post a brand new thread with your issue along with a full syslog and one of us will be able to walk you through the issue you have due to your motherboard. In the meantime do a search on HPA in the wiki or on the forum to understand what your issue most likely is. Quote Link to comment
Chris Pollard Posted February 8, 2011 Share Posted February 8, 2011 Post a fix subject title in the forum that THE program can fix this problem and I will try again. Have you actually read up on the background, and understood the root of this problem? Your motherboard makes a hidden reservation of a chunk of the hard drive and you expect the application to retrieve the space??? To be fair, it worked fine until now and gigabyte mobo's aren't exactly uncommon. Quote Link to comment
BRiT Posted February 8, 2011 Share Posted February 8, 2011 And to be fair, there are all sorts of warnings in the forums and wiki about HPA and Gigabyte motherboards. Quote Link to comment
prostuff1 Posted February 8, 2011 Share Posted February 8, 2011 To be fair, it worked fine until now and gigabyte mobo's aren't exactly uncommon. It may have "worked fine" but I think there is some other stuff going on here. A switch in the kernel or something of the like that is detecting the HPA and showing this issue. This switch may have had to be turned on/off because of support for drives larger than 2TB. It is not overly difficult to remove the HPA and get the full drive size back. Hex is over reacting a little bit as this is not an unRAID problem nor one that unRAID can solve by itself. His motherboard created the problem and to clear it up requires some user input. Quote Link to comment
limetech Posted February 8, 2011 Author Share Posted February 8, 2011 Same ol story. Updated to 4.7, samsung drive reports too small. Went back to 4.6 and am going to wait until the expensive program I bought can fix the problem. I am tired of patching program errors. Did I buy into a hobby program or a professional program? I am glad the program works good as is and I guess I will not update. Post a fix subject title in the forum that THE program can fix this problem and I will try again. Thanks I understand your frustration and I'm looking into an alternative way to handle same-model drives which have different reported size because of existence of HPA. Quote Link to comment
T3nchu Posted February 8, 2011 Share Posted February 8, 2011 Anyone got recommendation in how to repair my harddrive that show less space then what it should be? Read a few post up please. Quote Link to comment
ohlwiler Posted February 8, 2011 Share Posted February 8, 2011 Use SeaTools and "Set Capacity to Max Native" http://seagate.custkb.com/seagate/crm/selfservice/search.jsp?DocId=201271&NewLang=en Be careful that your motherboard (Gigabyte) does not create a new HPA. Quote Link to comment
T3nchu Posted February 9, 2011 Share Posted February 9, 2011 Use SeaTools and "Set Capacity to Max Native" http://seagate.custkb.com/seagate/crm/selfservice/search.jsp?DocId=201271&NewLang=en Be careful that your motherboard (Gigabyte) does not create a new HPA. I don't have any PC here I did boot up with my Mac with bootcamp but seatools get stuck on searching for devices...... DOS version does not recognize the USB mouse and keyboard. I also tried HDD Factory Capacity Restore, it gave a an error about not able to restore the device. I am in the process of replacing my MoBo, so I am not going to reboot until I get that new MoBo. I am not sure if I can recover the data on that drive now. Unfortunately, I did initconfig and resync the parity of the drive I had working (did not want to lost what I still have). 4.7 update killed me Is there any other way in regaining the full drive capacity, since I can't really save the data on the drive any more? Does simple reformat the drive work? Quote Link to comment
dgaschk Posted February 9, 2011 Share Posted February 9, 2011 Same ol story. Updated to 4.7, samsung drive reports too small. Went back to 4.6 and am going to wait until the expensive program I bought can fix the problem. I am tired of patching program errors. Did I buy into a hobby program or a professional program? I am glad the program works good as is and I guess I will not update. Post a fix subject title in the forum that THE program can fix this problem and I will try again. Thanks I understand your frustration and I'm looking into an alternative way to handle same-model drives which have different reported size because of existence of HPA. You shouldn't blame Limetech because your drives have HPA on them. If you were cognizant of the condition of your drives you wouldn't have this problem. Please accept responsibility for the condition of your drives; since it was your equipment that fouled up your drive in the first place. There are a lot of postings regarding this issue and solutions that do not require the purchase of any software have been detailed. Quote Link to comment
limetech Posted February 9, 2011 Author Share Posted February 9, 2011 Same ol story. Updated to 4.7, samsung drive reports too small. Went back to 4.6 and am going to wait until the expensive program I bought can fix the problem. I am tired of patching program errors. Did I buy into a hobby program or a professional program? I am glad the program works good as is and I guess I will not update. Post a fix subject title in the forum that THE program can fix this problem and I will try again. Thanks I understand your frustration and I'm looking into an alternative way to handle same-model drives which have different reported size because of existence of HPA. You shouldn't blame Limetech because your drives have HPA on them. If you were cognizant of the condition of your drives you wouldn't have this problem. Please accept responsibility for the condition of your drives; since it was your equipment that fouled up your drive in the first place. There are a lot of postings regarding this issue and solutions that do not require the purchase of any software have been detailed. Actually I should have come up with a better solution so that existing already-formatted drives get recognized correctly - I didn't anticipate an HPA that uses an odd number of sectors, having never run across one in any server, though of course they are obviously out there. Edit: I should add that this issue can be corrected, but there is a general problem with HPA... suppose you are using say a parity disk and a couple data disks of all the same model and all with an HPA (so they are all smaller than raw capacity). Now suppose a disk fails so you replace with same model drive but unknown to you, it has no HPA. Now this drive will be bigger than all his brothers and this will be confusing and people will start posting: "I put in the same model drive but it's not the same size - what's the matter with the s/w?" There are ways of solving this but all are ugly. Complicating the matter is that for some reason, not all drivers are able to access the HPA (or rather not all drivers implement the ATA command to access the HPA). Ugh, I should have listened to my father and became a dentist. Quote Link to comment
dgaschk Posted February 9, 2011 Share Posted February 9, 2011 Same ol story. Updated to 4.7, samsung drive reports too small. Went back to 4.6 and am going to wait until the expensive program I bought can fix the problem. I am tired of patching program errors. Did I buy into a hobby program or a professional program? I am glad the program works good as is and I guess I will not update. Post a fix subject title in the forum that THE program can fix this problem and I will try again. Thanks I understand your frustration and I'm looking into an alternative way to handle same-model drives which have different reported size because of existence of HPA. You shouldn't blame Limetech because your drives have HPA on them. If you were cognizant of the condition of your drives you wouldn't have this problem. Please accept responsibility for the condition of your drives; since it was your equipment that fouled up your drive in the first place. There are a lot of postings regarding this issue and solutions that do not require the purchase of any software have been detailed. Actually I should have come up with a better solution so that existing already-formatted drives get recognized correctly - I didn't anticipate an HPA that uses an odd number of sectors, having never run across one in any server, though of course they are obviously out there. Edit: I should add that this issue can be corrected, but there is a general problem with HPA... suppose you are using say a parity disk and a couple data disks of all the same model and all with an HPA (so they are all smaller than raw capacity). Now suppose a disk fails so you replace with same model drive but unknown to you, it has no HPA. Now this drive will be bigger than all his brothers and this will be confusing and people will start posting: "I put in the same model drive but it's not the same size - what's the matter with the s/w?" There are ways of solving this but all are ugly. Complicating the matter is that for some reason, not all drivers are able to access the HPA (or rather not all drivers implement the ATA command to access the HPA). Ugh, I should have listened to my father and became a dentist. Please don't take time away from 5.0 development to fix what is basically a user error (even if based on common naivety). I asked Raj to pin an HPA thread at the top of the HD forum. We will detail solutions for various drives in the first post. Once it is complete, people can refer to it for a fix for any drive. If someone has HPA on all of their drives they will have to replace parity first or do a parity swap. Edit: perhaps when someone replaces a data drive with a drive that is larger than parity the UI could indicate that a parity swap is required. Quote Link to comment
Chris Pollard Posted February 9, 2011 Share Posted February 9, 2011 Its perfectly possible to end up with a drive with a HPA on it having never owned a Gigabyte motherboard. The first time it would affect you would be upgrading to 4.7. I see where people are coming from but at the end of the day this change in how things work has caused issues for a number of people now it seems so I at least think it would be prudent to make mention of the issue in the release notes. I've personally held off upgrading to this release because I can't be bothered with the hassle of messing about with the HPA on one of my drives (created by a motherboard I no longer own!) Quote Link to comment
jbrodriguez Posted February 10, 2011 Share Posted February 10, 2011 i succesfully upgraded to 4.7. before the upgrade, i had to remove hpa for 3 disks after the upgrade, i precleared an ears (without jumper, with the -A option) and was able to add it to the array without problems so far everything is working fine. Quote Link to comment
PeterB Posted February 11, 2011 Share Posted February 11, 2011 Its perfectly possible to end up with a drive with a HPA on it having never owned a Gigabyte motherboard. I guess that might be true ... if you buy a used drive from someone who does have a Gigabyte mobo! Quote Link to comment
CHBMB Posted February 17, 2011 Share Posted February 17, 2011 Same ol story. Updated to 4.7, samsung drive reports too small. Went back to 4.6 and am going to wait until the expensive program I bought can fix the problem. I am tired of patching program errors. Did I buy into a hobby program or a professional program? I am glad the program works good as is and I guess I will not update. Post a fix subject title in the forum that THE program can fix this problem and I will try again. Thanks Try WHS, I've got a copy you can have for 1/2 price! Complete with all manner of problems and workarounds, but with a lovely looking interface, of course the support is nowhere near as good as here and the updates are few and far between, but the new version (Vail) removes the most useful features and thereby removing the bugs. Honestly I can't fault the support here and the frequent updates and fantastic user community. Hands up those of us that have had a BSOD with a MS product, do you start accusing Mr Gates of producing a "Hobby program" I don't see how Tom can be held responsible for a problem created by a hardware manufacturer who forces functionality and alters a drive characteristics whether you want it to or not. Quote Link to comment
Hex Posted February 18, 2011 Share Posted February 18, 2011 I run a 10TB WHS V1 and it runs fine since I put all the drives on a sata port multiplier and got all the USB drives off of it. I also run a Windows Web server with a dozen sites on it. I have 2 FTP sites and also run my own email server. I am a programmer. I bought Unraid because of the unique way of data protection. It is not the double the data on two drives like WHS. I now have a 12TB Unraid that I cannot update because of a problem. I bought a motherboard that was on the approved list that didn't work. The next time I checked the list the MB was blacklisted. It was a rough road to get this thing to work but it does work. I have my unit as a headless unit stuck in a closet across the house. I guess I will telnet into it and fix it. I wrote that last message after a long day and I wasn't in any mood to modify parameters or track down logs or any of that stuff. But I guess I will do that now that I am in a better mood. Lime, if during the initial setup, can you just take the size of the disc as gospel or will bios choke on it? Y'know, eliminate the HPA during format? Don't stop work on 5+ or dgaschk will blow a vein. Thanks for your support Chris. Quote Link to comment
Joe L. Posted February 18, 2011 Share Posted February 18, 2011 The HPA is not something you can eliminate during the "formatting" of a disk. Boy, do I wish it was as easy as that. It is a feature built into the disk, one that gets to be set once per power cycle of a hard disk (as far as I've read) and one that some BIOS seem to have really gotten wrong. (Specifically Gigabyte on some MB) With all your experience, you'll understand nobody on these forums work for lime-tech other than "limetech" Good luck with your server. Joe L. Quote Link to comment
MartinQ Posted February 19, 2011 Share Posted February 19, 2011 Was just going to upgrade my parity drive from 1.5T to 2.0T with advanced-format so I figured I'd upgrade from 4.6 to 4.7 at the same time. 1. Did a parity check. OK 2. Shutdown server. 3. Swap drives. 4. Boot & check status. Parity drives shows up OK with blue-dot. 5. Copy over new bzimage & bzroot (overwrite 4.6 with 4.7) and reboot and now I've got four drives in the red! Feb 19 02:29:07 Watchtower kernel: md: unRAID driver 1.1.1 installed Feb 19 02:29:07 Watchtower kernel: md: import disk0: [8,128] (sdi) ST2000DL003-9VT1 5YD1DLK4 size: 1953514552 Feb 19 02:29:07 Watchtower kernel: md: disk0 wrong Feb 19 02:29:07 Watchtower kernel: md: import disk1: [8,32] (sdc) WDC WD5000AAKS-6 WD-WCAS85494000 size: 488386552 Feb 19 02:29:07 Watchtower kernel: md: import disk2: [8,48] (sdd) WDC WD5000AAKS-6 WD-WCAS85394666 size: 488386552 Feb 19 02:29:07 Watchtower kernel: md: import disk3: [8,16] (sdb) WDC WD5000AAKS-2 WD-WCAPW4129766 size: 488386552 Feb 19 02:29:07 Watchtower kernel: md: import disk4: [8,0] (sda) ST31500541AS 9XW02NFA size: 1465137492 Feb 19 02:29:07 Watchtower kernel: md: disk4 wrong Feb 19 02:29:07 Watchtower kernel: md: import disk5: [8,96] (sdg) WDC WD15EADS-00P WD-WCAVU0390002 size: 1465138552 Feb 19 02:29:07 Watchtower kernel: md: import disk6: [8,64] (sde) WDC WD5000AAKS-2 WD-WCAPW4107675 size: 488385492 Feb 19 02:29:07 Watchtower kernel: md: disk6 wrong Feb 19 02:29:07 Watchtower kernel: md: import disk7: [8,80] (sdf) WDC WD5000AAKS-2 WD-WCAPW4125131 size: 488385492 Feb 19 02:29:07 Watchtower kernel: md: disk7 wrong Feb 19 02:29:07 Watchtower kernel: md: import disk8: [8,112] (sdh) WDC WD5000AAKS-6 WD-WCAS85486418 size: 488386552 Quote Link to comment
MartinQ Posted February 19, 2011 Share Posted February 19, 2011 I just reverted back to 4.6 and got the blue dot back. Going to build parity on it and see how it goes. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.