June 13, 201511 yr A question regarding 'current pending sectors': Some history first... I am in the process or building a new Unraid 6 server, and am reappropriating disks from my Unraid 5 build, one of which being a WD20EARS drive. Previously I had the drive jumpered and set to MBR unaligned. With the new build, I have removed the jumper and reformatted in XFS. I did not do another pre-clear. Now I am in the process of rsyncing all my data back on to the server, and playing around with the nice new Unraid 6 UI, I noticed one line in orange: Current Pending Sector (see attached screenshot). Any ideas if this is actually a problem or if there is anything I should do? Cheers!
June 13, 201511 yr It is a problem. See here: http://lime-technology.com/wiki/index.php/Troubleshooting#Resolving_a_Pending_Sector
June 14, 201511 yr Author Thanks for the info. So I basically need to run a pre-clear cycle on it, and hopefully it will be resolved. Brilliant, I just finished restoring 2TB of data on to it. Does anyone know if this is anyway connected to the removal of the jumper?
June 14, 201511 yr Community Expert Thanks for the info. So I basically need to run a pre-clear cycle on it, and hopefully it will be resolved. Brilliant, I just finished restoring 2TB of data on to it. A pre-clear is the easiest way to do this (assuming that you have a backup of the data). In addition to hopefully clearing the current pending sectors it also does a confidence test on the rest of the disk. I know that if it is a Seagate drive then Seatools can often clear pending sectors without wiping the current contents. Other manufacturer's tools may offer similar capabilities. However whether data corruption can result for the 'cleared' sectors I am not sure so this would not be a way I would recommend anyone proceeding. Does anyone know if this is anyway connected to the removal of the jumper? I would think it was very unlikely.
June 14, 201511 yr Author Am I reading this right, there are >65k current pending sectors!?!?! (see screenshot in post 1)
June 14, 201511 yr Interesting. I suspect that's an erroneous value ... 65535 is -1 in a signed 16-bit register => and may very well be somehow related to removal of the jumper (although it's not clear how this would happen). You don't just need to reformat the drive; you need to re-initialize it. I'd attach it to a PC and run WD's Data Lifeguard against it, doing a Write Zeroes to it with the FULL option. Then see if the S.M.A.R.T. data still shows a pending sector.
June 14, 201511 yr Author Thanks for the prompt reply. Out of interest, how does the WD tool differ from a pre-clear - I thought that went through writing 0's to the drive as well? Cheers.
June 14, 201511 yr A pre-clear should do it as well ... I just prefer the manufacturer's diagnostic utility, which will also offer to repair damaged sectors if they're repairable. I'd do a full write zeroes, followed by the extended test. The two together take ~ half as long as a pre-clear cycle.
June 14, 201511 yr Thanks for the info. So I basically need to run a pre-clear cycle on it, and hopefully it will be resolved. Brilliant, I just finished restoring 2TB of data on to it. Does anyone know if this is anyway connected to the removal of the jumper? Some have reported in the past some strange issues removing jumper after have used WDs with jumper for a while... Anyway, you can try with a HDD Regenerator 'Scan and repair' surface scan which doesn't require to cancel data from hard drive for the process...
June 14, 201511 yr Author I can always stick the jumper back in and then set it back to MBR unaligned, or is that not going to help at all?
June 16, 201511 yr Author Ok, drive cleared - COMPLETED. Extended test run - PASSED. End SMART test still showing 1 current pending sector. Doesn't the powerdown plugin write SMART reports to a log file on shutting down? If so I can probably dig them up from my Unraid 5 server build, maybe that would then at least show if the current pending sector was there before?!? Next steps - I can try replacing the jumper, I could live with it I guess... Thanks for the help guys.
June 16, 201511 yr The "Value" column you highlighted is NOT the # of pending sectors. That column is the value associated with each of the parameters. Data Lifeguard doesn't show the actual raw values, which is what you want to see now. If the drive is still attached to a Windows machine, download PassMark's free Disk Checkup and look at the SMART values with it. http://www.passmark.com/products/diskcheckup.htm BTW, There's no reason to replace the jumper.
June 16, 201511 yr Author The "Value" column you highlighted is NOT the # of pending sectors. That column is the value associated with each of the parameters. Man they don't make this easy do they?! Will check it again tonight using Passmark. Cheers.
June 16, 201511 yr I don't know why WD doesn't show the raw values in the SMART display in Data Lifeguard, but for whatever reason they choose not to. The "Value" column you're looking at is actually a "Normalized Value" that is set differently by different manufacturers. Typically a "perfect" setting for this is either 100, 200, or 253 => and as it declines it's getting worse. A value of 1 is very low, but it's still above the threshold value of 0, so it's not a "failing" parameter. What you need to see is the current raw value, which WAS 65535 earlier. If that's now cleared all is fine. Actually, even if it's not, the drive is likely fine ... it's likely a "glitch" in the SMART data ... but I'd still take it out of service in the array, just to be safe [use it as a backup drive]. Note that if you disable spindown (at least for the troublesome drive) and let it "idle" for a day or so, it MAY self-correct the pending count. Most modern drives will do an automatic "off-line scan" when they're idle for an extended time to check certain SMART attributes ... and the pending sectors is one of those areas checked.
June 16, 201511 yr Author Ok, results from Passmark attached... Server is literally off most of the time - I only spin it up to make backups, watch a film, listen to music etc... So I can certainly try re-installing it in the array, turn off spin down and see how things are in 24 hours. BTW, what's the best way to go about reinserting this, I guess I let Unraid do a data recreate upon starting the array??
June 16, 201511 yr BTW, what's the best way to go about reinserting this, I guess I let Unraid do a data recreate upon starting the array?? How did you remove it? Is it now a "missing" disk in the array? Or did you do a New Config so it's not part of the array? If it's a "missing" disk you can put it back; re-assign it to the slow with the missing disk; and then UnRAID will do a data rebuild. If it's not currently part of the array, you can add it to the array ... but there won't be any data rebuilt onto the disk.
June 16, 201511 yr Author It will be missing I guess (I have yet to turn the server back on). So yes, I will do a data rebuild. Here are the exported values from Disk Checkup: *** DiskCheckup V3.3 Build: 1000 Report *** SysInfo DLL Version: SysInfo v1.0 Build: 1090 Time of export: 18:17:57 16-Jun-2015 Device information: Device ID: 1 Interface: USB Device Capacity: 1907726 MB Serial Number: WD-WMAZA1718008 Model Number: WDC WD20EARS-00MVWB0 SMART ATTRIBUTES: ID Description Status Value Worst Threshold Raw Value TEC --------------------------------------------------------------------------------------------------------------------------------------------- 1 Raw Read Error Rate OK 198 197 51 1447 N.A. 3 Spin Up Time OK 242 169 21 2858 N.A. 4 Start/Stop Count OK 99 99 0 1126 N.A. 5 Reallocated Sector Count OK 200 200 140 0 N.A. 7 Seek Error Rate OK 200 200 0 0 N.A. 9 Power On Time OK 97 97 0 2432 N.A. 10 Spin Retry Count OK 100 100 0 0 N.A. 11 Calibration Retry Count OK 100 100 0 0 N.A. 12 Power Cycle Count OK 99 99 0 1000 N.A. 192 Power off Retract Count OK 200 200 0 18 N.A. 193 Load Cycle Count OK 179 179 0 63367 N.A. 194 Temperature OK 119 107 0 31 C N.A. 196 Reallocation Event Count OK 200 200 0 0 N.A. 197 Current Pending Sector Count OK 1 1 0 65535 N.A. 198 Uncorrectable Sector Count OK 200 200 0 0 N.A. 199 UltraDMA CRC Error Count OK 200 200 0 4 N.A. 200 Write Error Count OK 199 199 0 373 N.A.
June 16, 201511 yr As I'm sure you noticed, the S.M.A.R.T. data still shows the (almost certainly incorrect) 65535 value for pending sectors. This is either a -0 or a -1 in a signed binary register (depends on whether the drive's firmware uses one's complement or two's complement. In either case, I don't think it's a valid number ... but not at all sure how (or if) it can be corrected. The drive is likely fine ... I doubt it'll cause any issues at all; but if you want to be overly cautious I'd just use it as a backup drive and replace it with a drive without that issue.
June 16, 201511 yr Author Hmm this is funny. I reinstalled the drive in its original slot. Started the server and it states 'disk format unknown' (or something to that effect). So of course I opted to format the drive (which took longer than expected TBH), and now the array is started and parity is valid. I would have expected to be asked if I want to recreate the drive from parity...
June 16, 201511 yr Community Expert Formatting a drive wipes its existing contents so I hope that is what you wanted to do.
June 16, 201511 yr Author Sure, well the drive was "zero'd" so it had no format. Since I'd already used the drive in the server I would have expected this to be recognised and then asked to recreate data on to it, once the drive had a valid disk format.
June 16, 201511 yr Community Expert Sure, well the drive was "zero'd" so it had no format. Since I'd already used the drive in the server I would have expected this to be recognised and then asked to recreate data on to it, once the drive had a valid disk format. That is unfortunately not the way rebuilding data works. The rebuild function works at the physical level and does not care about disk format or whether a disk is formatted at all. Formatting a disk tells unRAID to throw away its existing contents and create a new empty directory structure on the disk.
June 16, 201511 yr Author Ok, that I understand. So in future when I re-insert a drive, which potentially won't have a file system, how do I get Unraid to rebuild from parity? As I looked at the various options and did not see something like "rebuild".
Archived
This topic is now archived and is closed to further replies.