March 4, 201313 yr OK - so I'm rebuilding my unRaid server into a new case and decided to add a 3TB Parity drive so that I can up all my 2TB's to 3TB's at some time in the future. Bought a WESTERN DIGITAL Caviar Green WD30EZRX 3.5" 3 TB (couldn't really afford a enterprize drive so got a "standard" one) and put it into the new case, put a minimum unRaid onto a memory stick (with pre-clear) and started a pre clear_disk.sh. Went away for the weekend assuming that it would be finished when I got back.... and maybe it had but all I could see on the screen was a load of numbers that kept repeating over and over again (I know - I should have written them down but I never and I know, I should have saved a log - but it never occurred to me). I assumed that maybe somebody had prodded a few keys on the keyboard and so escaped and did shutdown - intending to start again with another pre-clear (but this time keep an eye on it!) Stared a preclear running again and this time it completed in 5 minutes?!!!!! I saved a log this time and got the attached log file. Could somebody who understands these things take a look at it and tell me whether it really has pre cleared and why it will not do it again?! Since it's a going to be a parity drive I want it to be OK... so any tips or advice really gratefully received. Thanks in advance - hope somebody can advise! DaveK preclear_results.txt
March 9, 201313 yr It did not clear... In fact, apparently the disk stopped responding after a few minutes. Perhaps a cable came loose? Joe L.
March 9, 201313 yr Author Thanks for this Joe L. - was about to give up hope of a response. I've tried a few more pre-clear attempts on the same drive (but this time I've kept more of an eye on it) and pretty much the same thing has happened. I get a "divide by zero" error from the preclear application too occasionally when it stops. I guess that it could be a dodgy connection (the drive is just dangling in the carcass of my new case whilst I test the new mobo out (eg IPMI etc) but that would probably also stop it from starting a preclear too I suppose? Wanted to preclear this brand new WD Green 3TB in preparation for moving my old unRaid machine contents to my new unRaid .... thought I could add 3TB Parity at the same time ... but having second thoughts now. Think I'll get my old unRaid system running in my new unRaid case/mobo and then perhaps try the preclear again with the 3TB in my old unRaid case.... this has hotswap bays and so will not have to dangle it. If this fails then its an RMA. If this is caused by a faulty drive... have you heard of this sort of issue in the past (ie preclear stops after a few minutes because the disk is faulty?) Thanks again DaveK
March 9, 201313 yr I have the exact same issue with my green 3TB drive, I fixed it (so it'd last longer) by running it through a long format on windows, however, now it just crashes after a couple of hours vs a couple of minutes. Subscribing to this thread, I want to see if/how you fix it, may work on mine.
March 9, 201313 yr I have the exact same issue with my green 3TB drive, I fixed it (so it'd last longer) by running it through a long format on windows, however, now it just crashes after a couple of hours vs a couple of minutes. Subscribing to this thread, I want to see if/how you fix it, may work on mine. Writing "zeros" to it on windows was not a fix. Windows probably just ignores most errors and certainly does not re-read the disk to ensure it was written correctly, or re-try if it fails.... If the disk is still failing now, and failed before, no actual fix occurred on windows. As you use the disk, it might "break in" and improve slightly as something binding might just eventually free itself, but your disk is not in great health. (and windows no great saviour) The big difference is windows does not use every sector of every disk and just crashes or corrupts your data when a sector is misread. The "fix" is replacement... (there is no other alternative unless you are using a bad or under-capacity power supply, in which case, the fix is still replacement, but of the power supply)
March 9, 201313 yr I have the exact same issue with my green 3TB drive, I fixed it (so it'd last longer) by running it through a long format on windows, however, now it just crashes after a couple of hours vs a couple of minutes. Subscribing to this thread, I want to see if/how you fix it, may work on mine. Writing "zeros" to it on windows was not a fix. Windows probably just ignores most errors and certainly does not re-read the disk to ensure it was written correctly, or re-try if it fails.... If the disk is still failing now, and failed before, no actual fix occurred on windows. As you use the disk, it might "break in" and improve slightly as something binding might just eventually free itself, but your disk is not in great health. (and windows no great saviour) The big difference is windows does not use every sector of every disk and just crashes or corrupts your data when a sector is misread. The "fix" is replacement... (there is no other alternative unless you are using a bad or under-capacity power supply, in which case, the fix is still replacement, but of the power supply) What I mean is, before the disk would fail right away with unraid, it's instantly crash and not respond to anything until I did a power cycle. I put it into windows, did a full format then put it into unraid again, this time it doesn't instantly fail and now fails nearer the end. Either way, I call that a fix of some degree, now it at-least is able to write/read the disk instead instead of instantly failing. Also, I will replace it, just not now. I'm going on holiday soon and I won't be able to send it to them/receive it, I'll RMA it the second I'm back, however. Warranty is till November 2014 I believe and I still have a few TB remaining on my server, along with a 1TB drive I've yet to assign. I'll live without it, however, I'm open to reading this thread and trying stuff, can't hurt, can it? If anything the mass amounts of trials against the drive will degrade it even more and when I finally do try and RMA it, it'll be even easier because not just the firmware is crashing (I believe?) but also the fact that the motor is failing/smart is failing.
March 19, 201313 yr Author UPDATE! Ok - finished the unRaid rebuild and so retried preclearing the 3TB WD Green. This time I ensured that it was plugged into the Mobo SATA port and not into the SAS/SATA PCI ports. I set it up to do 3 passes - and this time it seemed to be working OK but stopped after 9 hours and created a 1.5gb log file. The screen was simply full of characters printing over and over again. I won't try to upload the log file because it just says the same thing over and over and over again: Mar 19 04:40:01 Tower syslogd 1.4.1: restart. Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] Unhandled error code Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] Result: hostbyte=0x04 driverbyte=0x00 Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] CDB: cdb[0]=0x88: 88 00 00 00 00 01 5d 50 a1 40 00 00 00 08 00 00 Mar 19 04:40:01 Tower kernel: end_request: I/O error, dev sda, sector 5860532544 Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] Unhandled error code Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] Result: hostbyte=0x04 driverbyte=0x00 Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] CDB: cdb[0]=0x88: 88 00 00 00 00 01 5d 50 a1 40 00 00 00 08 00 00 Any input anyone ... have I got a bad drive or could it be something else. I don't think its a bad connection as it is a brand new build and I've checked out all the connections carefully. Help?! Please?! Thanks in anticipation DaveK
March 19, 201313 yr UPDATE! Ok - finished the unRaid rebuild and so retried preclearing the 3TB WD Green. This time I ensured that it was plugged into the Mobo SATA port and not into the SAS/SATA PCI ports. I set it up to do 3 passes - and this time it seemed to be working OK but stopped after 9 hours and created a 1.5gb log file. The screen was simply full of characters printing over and over again. I won't try to upload the log file because it just says the same thing over and over and over again: Mar 19 04:40:01 Tower syslogd 1.4.1: restart. Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] Unhandled error code Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] Result: hostbyte=0x04 driverbyte=0x00 Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] CDB: cdb[0]=0x88: 88 00 00 00 00 01 5d 50 a1 40 00 00 00 08 00 00 Mar 19 04:40:01 Tower kernel: end_request: I/O error, dev sda, sector 5860532544 Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] Unhandled error code Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] Result: hostbyte=0x04 driverbyte=0x00 Mar 19 04:40:01 Tower kernel: sd 2:0:0:0: [sda] CDB: cdb[0]=0x88: 88 00 00 00 00 01 5d 50 a1 40 00 00 00 08 00 00 Any input anyone ... have I got a bad drive or could it be something else. I don't think its a bad connection as it is a brand new build and I've checked out all the connections carefully. Help?! Please?! Thanks in anticipation DaveK You probably want to have a look at my thread, identical log (I believe, I can't remember off the top of my head, looks close enough) and everyone pretty much just told me to RMA it. @Joe, before you come in telling me that's what I should do, I will, once I have time. I'm not in dire need of space at the moment and I have awhile left on my warranty.
March 19, 201313 yr Author Hi! Thanks for getting back to me so quickly - OK - will RMA it and hope that the replacement behaves better! Any chance you can point me at your thread that covers the same error?! Thanks again
March 19, 201313 yr Hi! Thanks for getting back to me so quickly - OK - will RMA it and hope that the replacement behaves better! Any chance you can point me at your thread that covers the same error?! Thanks again http://lime-technology.com/forum/index.php?topic=26276
April 11, 201313 yr Author Update - solved! So - I finally got a replacement "recertified" 3TB from Western after RMA's the drive that I described above. Just done 3 successful PreClears with 0 errors/problems reported. Took nearly a week - but probably worth it for the peace of mind. However I now have another question - but will post in a new topic Thanks everyone for your advice - really helped in solving the problem and also in increasing my unRaid knowledge a little too DaveK
April 12, 201313 yr Update - solved!(?? ?? ???) So - I finally got a replacement "recertified" 3TB from Western after RMA's the drive that I described above. Just done 3 successful PreClears with 0 errors/problems reported. Took nearly a week - but probably worth it for the peace of mind. However I now have another question - but will post in a new topic Thanks everyone for your advice - really helped in solving the problem and also in increasing my unRaid knowledge a little too DaveK That's fine.I had the same problem and it solved too.
Archived
This topic is now archived and is closed to further replies.