October 2, 201312 yr Just completed my monthly check. it completed, and says 'parity is valide' but also says Last checked on Tue Oct 1 19:23:33 2013 CDT, finding 7961 errors Never found errors before... i did have a hard reboot during the month, and also upgraded to 5.0. i think the monthly check runs with 'nocorrect' which hasn't been a problem until now... what should i do? shall i correct the errors? if i should correct them, how do i do that? any help appreciated.
October 2, 201312 yr Run another non-correcting check, and compare the total number and locations of the errors. Rough comparison, I don't mean you need to verify them one for one. If the errors are the same, and none of your drives show abnormal smart reports and the syslog looks clean, then I'd chalk it up to "one of those things", run a correcting check, another non-correcting check to make sure it took, and keep an eye on it. When you did the hard shutdown, or at any time, did you happen to cancel a parity check during this time period? Does your syslog show any drive errors?
October 2, 201312 yr Author Ok thanks will do this. Where do i find info on the parity check errors? In the sys log? It is possible that a hard shutdown occured during a parity check but not positive ( not good i know). My parity checks take about 20 hours to run so this may tske a while
October 2, 201312 yr Author here is my syslog. still trying to find out how to get smart reports. the errors all look clustered around one location but not sure what its telling me... syslog.zip
October 3, 201312 yr ~/Downloads/syslog-24q.txt:4764: Oct 1 09:58:17 Tower kernel: md: parity incorrect, sector=2528152672 ~/Downloads/syslog-24q.txt:4778: Oct 1 10:30:35 Tower kernel: md: parity incorrect, sector=2685141040 ~/Downloads/syslog-24q.txt:5006: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220208 ~/Downloads/syslog-24q.txt:5007: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220216 ~/Downloads/syslog-24q.txt:5008: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220224 ~/Downloads/syslog-24q.txt:5009: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220232 ~/Downloads/syslog-24q.txt:5010: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220240 ~/Downloads/syslog-24q.txt:5011: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220248 ~/Downloads/syslog-24q.txt:5012: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220256 ~/Downloads/syslog-24q.txt:5013: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220264 ~/Downloads/syslog-24q.txt:5014: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220272 ~/Downloads/syslog-24q.txt:5015: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220280 ~/Downloads/syslog-24q.txt:5016: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220288 ~/Downloads/syslog-24q.txt:5017: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220296 ~/Downloads/syslog-24q.txt:5018: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220304 ~/Downloads/syslog-24q.txt:5019: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220312 ~/Downloads/syslog-24q.txt:5020: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220320 ~/Downloads/syslog-24q.txt:5021: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220328 ~/Downloads/syslog-24q.txt:5022: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220336 ~/Downloads/syslog-24q.txt:5023: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220344 ~/Downloads/syslog-24q.txt:5024: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220352 ~/Downloads/syslog-24q.txt:5025: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220360 ~/Downloads/syslog-24q.txt:5026: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220368 ~/Downloads/syslog-24q.txt:5027: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220376 ~/Downloads/syslog-24q.txt:5028: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220384 ~/Downloads/syslog-24q.txt:5029: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220392 ~/Downloads/syslog-24q.txt:5030: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220400 ~/Downloads/syslog-24q.txt:5031: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220408 ~/Downloads/syslog-24q.txt:5032: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220416 ~/Downloads/syslog-24q.txt:5033: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220424 ~/Downloads/syslog-24q.txt:5034: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220432 ~/Downloads/syslog-24q.txt:5035: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220440 ~/Downloads/syslog-24q.txt:5036: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220448 ~/Downloads/syslog-24q.txt:5037: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220456 ~/Downloads/syslog-24q.txt:5038: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220464 ~/Downloads/syslog-24q.txt:5039: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220472 ~/Downloads/syslog-24q.txt:5040: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220480 ~/Downloads/syslog-24q.txt:5041: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220488 ~/Downloads/syslog-24q.txt:5042: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220496 ~/Downloads/syslog-24q.txt:5043: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220504 ~/Downloads/syslog-24q.txt:5044: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220512 ~/Downloads/syslog-24q.txt:5045: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220520 ~/Downloads/syslog-24q.txt:5046: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220528 ~/Downloads/syslog-24q.txt:5047: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220536 ~/Downloads/syslog-24q.txt:5048: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220544 ~/Downloads/syslog-24q.txt:5049: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220552 ~/Downloads/syslog-24q.txt:5050: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220560 ~/Downloads/syslog-24q.txt:5051: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220568 ~/Downloads/syslog-24q.txt:5052: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220576 ~/Downloads/syslog-24q.txt:5053: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220584 ~/Downloads/syslog-24q.txt:5054: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220592 ~/Downloads/syslog-24q.txt:5055: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220600 ~/Downloads/syslog-24q.txt:5056: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220608 ~/Downloads/syslog-24q.txt:5057: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220616 ~/Downloads/syslog-24q.txt:5058: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220624 ~/Downloads/syslog-24q.txt:5059: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220632 ~/Downloads/syslog-24q.txt:5060: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220640 ~/Downloads/syslog-24q.txt:5061: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220648 ~/Downloads/syslog-24q.txt:5062: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220656 ~/Downloads/syslog-24q.txt:5063: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220664 ~/Downloads/syslog-24q.txt:5064: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220672 ~/Downloads/syslog-24q.txt:5065: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220680 ~/Downloads/syslog-24q.txt:5066: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220688 ~/Downloads/syslog-24q.txt:5067: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220696 ~/Downloads/syslog-24q.txt:5068: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220704 ~/Downloads/syslog-24q.txt:5069: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220712 ~/Downloads/syslog-24q.txt:5070: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220720 ~/Downloads/syslog-24q.txt:5071: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220728 ~/Downloads/syslog-24q.txt:5072: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220736 ~/Downloads/syslog-24q.txt:5073: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220744 ~/Downloads/syslog-24q.txt:5074: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220752 ~/Downloads/syslog-24q.txt:5075: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220760 ~/Downloads/syslog-24q.txt:5076: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220768 ~/Downloads/syslog-24q.txt:5077: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220776 ~/Downloads/syslog-24q.txt:5078: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220784 ~/Downloads/syslog-24q.txt:5079: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220792 ~/Downloads/syslog-24q.txt:5080: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220800 ~/Downloads/syslog-24q.txt:5081: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220808 ~/Downloads/syslog-24q.txt:5082: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220816 ~/Downloads/syslog-24q.txt:5083: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220824 ~/Downloads/syslog-24q.txt:5084: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220832 ~/Downloads/syslog-24q.txt:5085: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220840 ~/Downloads/syslog-24q.txt:5086: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220848 ~/Downloads/syslog-24q.txt:5087: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220856 ~/Downloads/syslog-24q.txt:5088: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220864 ~/Downloads/syslog-24q.txt:5089: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220872 ~/Downloads/syslog-24q.txt:5090: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220880 ~/Downloads/syslog-24q.txt:5091: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220888 ~/Downloads/syslog-24q.txt:5092: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220896 ~/Downloads/syslog-24q.txt:5093: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220904 ~/Downloads/syslog-24q.txt:5094: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220912 ~/Downloads/syslog-24q.txt:5095: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220920 ~/Downloads/syslog-24q.txt:5096: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220928 ~/Downloads/syslog-24q.txt:5097: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220936 ~/Downloads/syslog-24q.txt:5098: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220944 ~/Downloads/syslog-24q.txt:5099: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220952 ~/Downloads/syslog-24q.txt:5100: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220960 ~/Downloads/syslog-24q.txt:5101: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220968 ~/Downloads/syslog-24q.txt:5102: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220976 ~/Downloads/syslog-24q.txt:5103: Oct 1 16:43:08 Tower kernel: md: parity incorrect, sector=4759220984 ~/Downloads/syslog-24q.txt:5104: Oct 1 16:43:08 Tower kernel: md: parity incorrect, stopped logging Run another check and compare these results. Do any drives show an error count on unRAID main?
October 3, 201312 yr Author ok will do. none of the drives show errors on unraid main ie. errors column has 0 for every drive
October 3, 201312 yr http://lime-technology.com/wiki/index.php/Troubleshooting#Obtaining_a_SMART_report post reports for all drives.
October 3, 201312 yr Author ok here are the reports! note that i ran these while i was doing my second nocorrect parity check as was suggested above! smart_reports.zip
October 3, 201312 yr Some quick comments before I hit the bed... I see 3 of your drives have pending sectors. WD-WMAZA0823880 (smarts.txt): 2 pending (0 reallocated) [LOAD CYCLES: 102214] WD-WCAWZ0251948 (smartr.txt): 41 pending (0 reallocated) [LOAD CYCLES: 81856] WD-WCAVY2471414 (smarti.txt): 10 pending (97 reallocated) [LOAD CYCLES: 84976] These are all WD greens with high load cycle counts as well. Other greens in the array have higher LCC'c than these. I would run the wdidle3 fix to mitigate the risk of them failing early... Someone more experienced than I should tell you how to proceed...
October 3, 201312 yr With no errors shown on any of the drives, it's almost certain that your parity errors are on the parity drive itself. The only surefire way to confirm this is to either verify checksums on the files on your array, or compare the files to your backups. If you have neither checksums nor backups, then I'd just assume the errors are on the parity drive, and run a correcting parity check. After that, run another check and confirm that there are no subsequent errors. As for the other two issues mentioned above ... (1) The load cycle counts are perhaps a bit high, although how high depends on how old the drives are. I looked at a couple of the drives -- one had 14,000 power-on hours, the other almost 20,000 hours. For drives that old, the load cycle counts aren't really bad at all. The drives are rated for 300,000 cycles ... and can often do several times that before encountering issues. But if you want to reduce the rate of increase, you can change the head parking counter to a longer time using the WDIDLE3 utility. The utility and instructions on its use are here: http://www.storagereview.com/how_to_stop_excessive_load_cycles_on_the_western_digital_2tb_caviar_green_wd20ears_with_wdidle3 (2) The pending reallocations. There's really nothing you need to do about these. The value for the parameter is still 200 in all cases (not decreased at all) => the pending sectors simply means the next time an attempt is made to write to that sector, it will be reallocated to a spare sector (decreasing the pending number and increasing the number of reallocations). You actually have NO reallocated sectors yet on any of the drives -- which means you haven't attempted a write to those sectors that have been flagged as "weak" (and marked for reallocation).
October 3, 201312 yr i personally would be concerned with any drive that has ANY pending sectors i think to start with i would address the disk that has reallocated sectors and still more pending sectors "WD-WCAVY2471414 (smarti.txt): 10 pending (97 reallocated) [LOAD CYCLES: 84976]" if the re-allocated sector count continues to rise i would def be looking at RMA that drive, as this indicates more sectors are failing on the drive. a section from the unraid wiki regards re-allocated sectors "Seeing a few reallocated sectors is not necessarily a bad thing, but seeing that number start to go up is often a sign that the drive is failing. Anytime you see a value other than 0 you should closely monitor the drive. If the number holds steady and does not increase even after several parity checks, your drive is likely okay. But if it seems to be going up by even 1 or 2 at a time, start to be concerned. This is likely the first hint that the drive is failing."
October 3, 201312 yr Reallocation of sectors is the way modern drives deal with defective sectors ... and while that count is perhaps a bit higher than I'd like to see, it's not all that high for a drive with over 21,000 hours on it; and the SMART value for pending reallocation is still 200 -- the same as it starts at, so at least SMART doesn't consider the counts excessive. I agree I'd track it to see if it's continuing to degrade. Before you decide you'd "... be looking at RMA that drive ..." you might want to see if that's an option The serial number is in the SMART report, so it's simple to pop over to WD's warranty check web site to see what the warranty status is => if you do so, you'll find that the warranty expired on 6 March 2013 -- about 7 months ago Whenever I have a drive with pending reallocations, I do a Level 3 pass with Spinrite on the drive -- but not everyone is willing to buy an $89 utility to do that with. That forces a complete refresh of the drive's surface -- rewriting every sector, which forces any pending reallocations to be done. But this isn't really necessary -- the drive will reallocate it the next time a write is attempted to that sector. The important thing isn't whether or not you have reallocated sectors (or pending ones) ... it's whether or not this count is consistently increasing => if so, you have a failing drive, and should consider replacing it.
October 3, 201312 yr Author Wow woke up to a lot of great feedback in this thread, i'm getting a good education and i'm grateful for all the well thought out responses! I will definitely keep an eye on those drives... running the smart reports wasn't hard at all (copied and pasted the script into putty, only changing the drive and filename each time)! Will specifically monitor the 3 drives mentioned for increasing pending counts. Thanks again. parity check still running, at 40%. no errors detected yet but will post back when completed
October 3, 201312 yr The LCC count will never cause a drive to fail SMART. The normalized VALUE may eventually reach 1 but it will never reach the TRESH. Values below 20 per hour are nothing to worry about. I've never seen or heard of a drive failing because of a high LLC count. Drive "s", for example, has a LCC count of over 100k and a VALUE of 166. 166 is not close to the THRESH of 0. The pending sectors are a serious issue. Unreadable sectors prevent unRAID from completing a rebuild of any other disk. If disk A has pending sectors then unRAID will fail when trying to rebuild a failed disk, B, C, or, D. If one drive has pending sectors the corrective action is to rebuild the disk with pending sectors. When multiple disks have pending sectors, they each prevent the rebuilding of any of the others. Copy the data from the disks with pending sectors. You may need to move the disk a into Windows machine and use utility software extract as much as possible. Then reset the array config without the faulty disks. Run pre-clear on the faulty disks to cause the pending sectors to be resolved. If the pending sector count goes to zero then add the drives back to the array.
October 3, 201312 yr Author yikes not good... not sure i totally understand- i think you're saying i need to copy data off the 3 disks, onto other disks within the array, then remove the drives to repair in my pc somehow? i'm not sure i even have room in the array to move 6tb into it! i do have a precleared 3tb drive (on the shelf in the case of needing a backup)... i could probably install the precleared drive, thereby making enough room on the array to copy off the 6tb, then remove the drives (and repair and or wipe them?) or maybe you're saying take them out now, so that parity and therefore protection can be restored immediately (minus the data on the 3 disks) then try to recover the data off the disks from my windows machine? again a problem because i don't have enough space on my windows machine to house the 6tb of data... sorry to be such a newb but maybe you could outline the steps you would take? i think it would make sense to add the extra drive i have into the array right away. then go into the disk shares (for the 3 affected drives) and copy that data into the non-affected drives (not sure how to do this, i use shares which span several drives, so not sure how to tell it to NOT use the affected drives). but if i could move the data over, then somehow rebuild parity using the array which no longer has the problem drives. then i can take time to fix/assess etc the affected drives (and wouldnt need to worry about recovering data because its already been copied over the array) oops edit: just realized i'm currently running the second parity check. should finish in a few hours, after which i was going to compare results to first check, and if similar, run a correcting check. maybe this is no longer required, as i think you're saying the array isn't protected anyway? GULP!
October 3, 201312 yr If you copy all the data from one of the disks; then write it back to it, that will clear the "pending" reallocations. Then just repeat that for each of the others. Note that if your data is important to you, you should have it backed up anyway -- so getting a few external drives to do this is a good idea
October 3, 201312 yr Author uggh sorry i'm feeling so slow today! are you saying to copy off the disk and then back on, but within the array? or copy off the affected disks, onto windows, then back again?
October 3, 201312 yr I'd copy it to a client PC (Windows for example) ... then copy it all back. Note that this will NOT absolutely clear the pending sectors -- if they're not written to, they won't be re-allocated. If you want to ensure that's done, you have to write to every sector on the disk. A fairly simple way (if a bit time-consuming) is to force the drive to be rebuilt. Stop the array; unassign the drive; Start the array (it should now show a "Missing" disk); Stop it again and re-assign the drive; then Start it again and it should rebuild the disk. That writes to every sector, so it will clear the "pendings". I simply do a Level 3 pass with Spinrite, which rewrites every sector on the drive and doesn't change anything (except, of course, pendings will be reallocated). But that utility costs $89.
October 4, 201312 yr I simply do a Level 3 pass with Spinrite, which rewrites every sector on the drive and doesn't change anything (except, of course, pendings will be reallocated). But that utility costs $89. $89 is a pretty good price for what it's capable of doing. I've brought more than 5 drives back to working order with it, and the upcoming 6.1 release will be much faster and work on more hardware (macs). If you have multiple hard drives it's a good utility to own. Sent from a mobile device, sorry for any typos.
October 4, 201312 yr Author ok so i forgot that i had a parity check running! if you recall i had 7961 errors, set to nocorrect. this came from my monthly ie October 1 parity check advice from the thread said to run it again and compare...attached is the new log but overall the errors seem very similar, although there is 7960 (one less than before) these are still NOT corrected... shall I correct them? i guess that will take another day or so to run... advice? syslog2.zip
October 4, 201312 yr I would correct them, but not all agree. Personally, I never run non-correcting checks. In ~ 5 years or so, I've only had 3 or 4 times when there were sync errors ... and just for grins, every time I spent a week or so running comparisons against my backups to ensure the errors had not been in the data => everything was perfect every time. About a year ago I started keeping checksums of all the data in addition to the backups ... so the next time I get a sync error (haven't had one since I started doing that) I can simply do a verify of the checksums ... won't even require getting out the backup disks.
October 4, 201312 yr Author Before it came up that i need to fix those three other drives i was planning on just correcting. Now im wondering if its just a waste of time since i guess the array is not really protected anyway. If i understood right since there are three drives with pending sector reallocations, if a drive dies tomorrow, the drive rebuild will fail. Not totally sure if im understanding correct though
October 4, 201312 yr Correcting parity is not possible if any drive cannot be correctly read. Attempt to copy from the drive to other working array drives. There will be problems copying files from the unreadable sectors. If files are missing then use linux or windows utility programs to recover as much as possible.
October 4, 201312 yr Author Ok i can try. I think i will have to add my precleared 3tb drive into the array and then copy one drives files on to it. I wonder what the best way to do that is? Set up a new share using only the brand new drive and then copy the entire disk share from the pending sector drive to this new share?
October 4, 201312 yr If none of your drives show read errors during a parity check, then I'd expect a correcting check to work just fine. A pending reallocation doesn't necessarily mean it can't be read -- it means it has had issues and is going to be reallocated on the next write to that sector ... but it may still be readable => and I'd expect that if there are no errors shown, that it was read just fine during the parity check [Doesn't, of course, mean that will always be the case]. Ideally, however, I agree it's best to clear the pendings before doing a correcting check => with Spinrite; or by copying all the data off and pre-clearing the drive (first removing it from the array); or any other method you may know. There's may be some Linux utility that will do this, but I'm not aware of it.
Archived
This topic is now archived and is closed to further replies.