June 20, 201313 yr Hi all and greetings from Syd Aus I just signed up and new to this whole thing unRAID in troubleshooting and im no Linux, command line expert for this software either (yes...scream, facepalm, run away if you wish). So apologies if this is in the wrong section or for being such a noob at this. Looking for some help and or any info. I just want to know whats wrong or if I should be looking elsewhere as I think its a drive problem but since im having some issues after restart (slow, unable to go through explorer or GUI) any feedback will do to get me started. I don't care the outcome if I have to rebuild this whole thing just want to know the problem, cause and what you all think. I'm running version 5.0-rc8a Brief spec of my system HDD- 3x Western Digital WD Green WD30EZRX 3TB SATA 6.0Gb/s MOBO - Asus P8H77-I H77 CPU- Intel Pentium G630 Processor - BX80623G630 RAM - Corsair XMS3 4GB (2x 2GB) DDR3 CMX4GX3M2B1600C9 CASE- Bitfenix prodigy PSU - Antec NEO ECO 400C 400W Power Supply (NEO-ECO-400C) About two months ago I had a hiccup with the parity drive being invalid (orange ball indicator). After a sync and check was fine until a few days ago...red ball (disabled). So I followed through the wiki in re-enabling it which worked. parity sync finished over night and seemed all ok (all 3 green balls) Now I ran a party check and its slow as hell some whopping 205-35 days for the check depending when I fresh the GUI of unRAID. most of the times now it is dead slow, sometimes cannot access unRAID menu/console and when I do its SO slow. When I try it won't prompt me to input the root password. I have checked the cabling and even replaced the SATA cable with a new one of the parity drive but did not try a different SATA port on the board. So I managed to do the smarttest command for the parity drive so I'll have to post some more info or a syslog (have to look for that command) but any help at all or suggestions would be great. Very late now and I'll have to leave it for the tomorrow and the weekend. if anyone can help me? I didn't expect this thing to fail in such a short time. I guess this is good that it happened now, its actually forcing me to read and learn about unRAID after building. to me it was get the parts, build, set it up, store crap, worry later. Thank you. Lnrds smarttest.zip
June 20, 201313 yr Looks to me like your drive is dieing and you need to replace it if you can. You have many relocated sectors and even worse many pending relocation. You need to at least run some pre-clear cycles until the pending sector count goes to zero. If it does and your relocated sector count does not increase any more you may be alright but personally I would replace it as soon as I could. If it is still under waranty then I personally would RMA it but as long as you have no pending realocation you might be alright for a while. Especially if the relocated count doesn't increase. If you choose not to RMA it then get periodic smart reports and make sure the pending and relocated counts don't increase. NOTE for proper operation pending should always be zero. This is where I was looking in your smart report for that info: 1 Raw_Read_Error_Rate 0x002f 200 200 051 Pre-fail Always - 8418 3 Spin_Up_Time 0x0027 193 175 021 Pre-fail Always - 5350 4 Start_Stop_Count 0x0032 100 100 000 Old_age Always - 109 5 Reallocated_Sector_Ct 0x0033 187 187 140 Pre-fail Always - 389 7 Seek_Error_Rate 0x002e 200 200 000 Old_age Always - 0 9 Power_On_Hours 0x0032 097 097 000 Old_age Always - 2409 10 Spin_Retry_Count 0x0032 100 100 000 Old_age Always - 0 11 Calibration_Retry_Count 0x0032 100 253 000 Old_age Always - 0 12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 37 192 Power-Off_Retract_Count 0x0032 200 200 000 Old_age Always - 16 193 Load_Cycle_Count 0x0032 200 200 000 Old_age Always - 2575 194 Temperature_Celsius 0x0022 125 109 000 Old_age Always - 25 196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 100 197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 110 198 Offline_Uncorrectable 0x0030 200 200 000 Old_age Offline - 328 199 UDMA_CRC_Error_Count 0x0032 200 200 000 Old_age Always - 0 200 Multi_Zone_Error_Rate 0x0008 199 199 000 Old_age Offline - 770 So get this line to zero instead of 110 at least: 197 Current_Pending_Sector 0x0032 200 200 000 Old_age Always - 110 Also the fact that you just had some smart ATA errors - report will only show last 5 but they all appear to have just happened - makes me think it is dieing as well. That info is after what I posted above and the power on hours 2409 above is the same as what is reported in the ATA errors.
June 20, 201313 yr Author Hi Bob Phoenix, Thank you very much sir. I suspected the drive is dud. Im starting to think and had discussed with my friend if its worth while using a Western Digital Red rather than a Green? Perhaps greens don't cope well for parity? Its possible I got the bad batch out of a billion hard drives coming out of Thailand lol. Unbelievable! such a short time but thank you for making the making the log clearer for me. Lnrds
June 20, 201313 yr Your drive definitely is bad/going bad => lots of pending sector reallocations; high error counts; etc. In my experience the WD Reds are MUCH better drives than the Green series. I've become a real "WD Red" zealot ... great performance; very low power consumption; and excellent reliability.
June 20, 201313 yr Author Hi garycase, Yes, I'm at work and currently going to buy another Red on my lunch break (4hrs). Looks like i'm going to have to pay to go pro when I get the green RMA'd back. I'd be able to fit in a total of 6-7 discs in my box. Do you recommend I should pre clear the Red or whack it straight in? takes forever
June 21, 201313 yr Do you recommend I should pre clear the Red or whack it straight in? takes foreverDo you feel lucky? Google drive failure bathtub curve. Preclear gives a measure of confidence that the drive isn't going to die an early death right after you get done trusting your data to it.
June 21, 201313 yr Normally I'd suggest you run a pre-clear first ... especially if this was a data drive. However ... considering that (a) you don't have good parity now, since your parity drive has failed; (b) the Reds are VERY reliable drives (I've VERY thoroughly tested the 15 I've purchased -- WD Lifeguard short, extended, zeroes, short, extended, followed by 2 preclear cycles ... about 60 hrs of testing/drive ==> and had ZERO failures); and © a parity sync followed by a parity test is a very thorough test by itself ... every sector will be written; then they'll all be checked ... ... I'd just put it in and let your array get some parity protection. After the parity sync, run a parity check; and then look at the SMART parameters on the drive. If the SMART looks good, then I'm sure you're fine.
June 21, 201313 yr Author Hi all, Thank you very much for your feedback I'll be getting the red soon and will post any issues I have. Again thank you all very much. Lnrds
June 21, 201313 yr Personally I would never add a drive that hasn't been throughly tested. I've had one WD Red DOA on me which is small compared to my failure rate on WD Greens but I would never use a drive without preclearing it first. If you are worried about being unprotected for a while then just turn off the box while you run some preclear cycles. I always do 3 cycles on a new drive and would never do less. I don't do more than 3 because like you I just can't wait more than the 3+ days it takes to do 3 cycles on a 3TB drive.
June 21, 201313 yr As I noted earlier, I agree. I ALWAYS pre-clear a DATA drive ... and would normally pre-clear a parity drive. But if you ARE going to "take a chance", the parity drive is the one drive that it's harmless to do so. If the drive fails, you don't have a parity drive. If you don't install it, you don't have a parity drive. And if the drive works well, your array is fault-tolerant 3 days earlier A full parity sync, followed by a parity check, is a pretty good exercise of the drive ... every sector written; every sector read; etc. If there are zero detected issues in SMART after that, it's a reasonable assumption that the drive's good. I should have mentioned one more thing: It'd be a good idea to do a SMART report BEFORE initiating the parity sync, and saving the results. Then compare that to a SMART report after the parity sync and parity checks. If there are no differences (other than hours, load cycles, etc. -- things that SHOULD change) then I'd be very confident in that drive. I agree, however, that if the array can simply be shut down for 3 days, then a pre-clear is a more thorough test.
Archived
This topic is now archived and is closed to further replies.