June 20, 201313 yr I'm getting allot of parity errors guess I had over 1700 at the beginning of the month and yesterday after running parity check I had over 200. Im running 5.0 RC15a with Unmenu 1.6. I try to get the smart history from unmenu but get the error "Sorry: cannot run smarthistory. The /boot/smarthistory directory does not exist." Any ideas as what could be going on with my system and how I can retrieve the smart history to get some more information posted about my issue? I was able to run a shot SMART report and attached it below smart.txt syslog-2013-06-23.txt
June 20, 201313 yr As I assume you know, if you run a correcting check and it shows 223 corrections (as it appears you did) ... then if you run another check immediately afterwards it SHOULD show zero corrections. When you ran the parity check, what did the "Errors" column show for each of the disks? ... and have you shut down the array and re-seated all of your SATA and data cables?
June 20, 201313 yr Smarthistory is an add-on package that can be invoked via unMENU link. If you did not install it, and not run it on a routine basis, there is no history to display. Apparently, since the directory it creates does not exist on your server, you've never installed it. Joe L.
June 20, 201313 yr Author I just ran a SMART report see above. I Just re-ran parity-check again and it started finding more errors after I just did a full one with correct-errors checked at that point I just canceled it. My errors column on all drives is zero on the main unraid screen though.
June 20, 201313 yr With no drives reporting errors, but a lot of parity sync issues, it sounds like you're having problems reading/writing to the parity drive (or perhaps it's failing). Did you try: (a) Re-seating the data and power cables to the parity drive? (b) Using a DIFFERENT SATA cable for that drive? ... also, do you by chance have your SATA cables tied together (bundled)?? If so, UNBUNDLE them => bundling a bunch of unshielded cables together can cause significant crosstalk when a lot of the attached drives are accessed simultaneously (as it clearly the case with a parity check).
June 20, 201313 yr Author With no drives reporting errors, but a lot of parity sync issues, it sounds like you're having problems reading/writing to the parity drive (or perhaps it's failing). Did you try: (a) Re-seating the data and power cables to the parity drive? (b) Using a DIFFERENT SATA cable for that drive? ... also, do you by chance have your SATA cables tied together (bundled)?? If so, UNBUNDLE them => bundling a bunch of unshielded cables together can cause significant crosstalk when a lot of the attached drives are accessed simultaneously (as it clearly the case with a parity check). I just ordered another HD. I wanted to expand anyway. I will swap it and run a couple parity checks and see if that resolves it and if so RMA the drive. Is a 6 month old WD RED drive.
June 22, 201313 yr Author New parity drive did not fix my issues started getting sync errors again after running parity check replaced the Sata cable on the parity drive and will do two more parity checks to see if error come up.
June 23, 201313 yr Author Run memtest. Bad memory is one of the more common parity error causes. Thanks for the tip. I will do that.
June 23, 201313 yr Author Let the memtest run overnight. The test did 18 passes and found no errors. I got the Parity drive in one of those cheap single hotswap bays I will remove it from that and just plug it straight in and see how it goes next. Is frustrating trying to track the issue down. Updated SYS LOG attached up top.
June 23, 201313 yr Jun 23 11:37:04 HAL kernel: md: import disk1: [3,0] (hda) ST32000542AS_6XW1RY37 size: 1953514552 Jun 23 11:37:04 HAL kernel: mdcmd (3): import 2 8,112 1953514552 WDC_WD20EARS-00MVWB0_WD-WCAZA1495454 Jun 23 11:37:04 HAL kernel: md: import disk2: [8,112] (sdh) WDC_WD20EARS-00MVWB0_WD-WCAZA1495454 size: 1953514552 Jun 23 11:37:04 HAL kernel: mdcmd (4): import 3 8,96 1953514552 WDC_WD20EARS-00MVWB0_WD-WCAZA4992418 Jun 23 11:37:04 HAL kernel: md: import disk3: [8,96] (sdg) WDC_WD20EARS-00MVWB0_WD-WCAZA4992418 size: 1953514552 Jun 23 11:37:04 HAL kernel: mdcmd (5): import 4 8,64 1953514552 WDC_WD20EARS-00MVWB0_WD-WMAZA3970271 Jun 23 11:37:04 HAL kernel: md: import disk4: [8,64] (sde) WDC_WD20EARS-00MVWB0_WD-WMAZA3970271 size: 1953514552 Jun 23 11:37:04 HAL kernel: mdcmd (6): import 5 3,64 1953514552 WDC_WD20EARS-00MVWB0_WD-WMAZA7126440 Jun 23 11:37:04 HAL kernel: md: import disk5: [3,64] (hdb) WDC_WD20EARS-00MVWB0_WD-WMAZA7126440 size: 1953514552 Several SATA drives are not configured as AHCI. All disk device labels should begin with s. Several begin with h, e.g. hda. Correct this in BIOS.
June 23, 201313 yr Author Here are the smart reports on my disks. I ran reiserfsck --check on all my disks and found no corruption. https://www.dropbox.com/sh/i2cvy78o2lcwrr2/20lx_vjsq9 . I also corrected my disks in the bios and set it to AHCI and the all report with an S at the beginning now. I notice this on disk6 199 UDMA_CRC_Error_Count 0x003e 200 191 000 Old_age Always - 3208 I will try new sata cable. Smart_Reports.zip
June 24, 201313 yr Author Disk 8 started showing errors on the main menu in Unraid after doing parity again. It was an old 160gb drive that had no data on it. I will remove it from the array and use the "New Config" Utility to rebuild parity tonight and see how it goes.
June 25, 201313 yr What PSU and how many drives total? You can isolate the problem drive by using New Config and only assigning 1/2 of the drives. If the problem is gone then the issue is with one of the missing drives. If the problem persists the unassign 1/2 of the drives so 1/4 are remaining. Repeat until the problem drive has been identified. Start by removing the oldest drives.
June 25, 201313 yr Author 8 drives thermaltake psu 600 watts. when i find the bad drive how do i replace it with all the parity errors?
June 25, 201313 yr Parity is already gone unless you changed the parity drive before using New Config and save a copy of the flash drive. Hopefully you will be able to copy the files off of the disk in unRAID or in a Windows PC.
June 25, 201313 yr 8 drives thermaltake psu 600 watts. when i find the bad drive how do i replace it with all the parity errors? What is the model number?
June 25, 201313 yr TR2-600NL2NH You are at the limit on this PSU. There have been some 2 port PCIe SATA cards that would return bad data during parity checks but otherwise function correctly. The problem only occurs when both drives are read concurrently.
June 25, 201313 yr Author the pcie card has been in for 2 years and never caused an issue perhaps it has something to do with my unraid version.
June 25, 201313 yr It's hard to say why it's occurring. I suggest that you set a new config including only drives attached to the MB and if that works continue searching until the problem card or drive is isolated.
Archived
This topic is now archived and is closed to further replies.