December 1, 201213 yr Hello - I need some help today ! I ran parity last night & it failed. Looks like my parity disk has been disabled. Here is an excerpt of my syslog - it is too big to attach (12MB). Dec 1 04:40:01 Tower syslogd 1.4.1: restart. Dec 1 04:40:04 Tower kernel: ata2: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 Dec 1 04:40:04 Tower kernel: ata2: status=0x41 { DriveReady Error } Dec 1 04:40:04 Tower kernel: ata2: error=0x04 { DriveStatusError } Dec 1 04:40:04 Tower kernel: ata2: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 Dec 1 04:40:04 Tower kernel: ata2: status=0x41 { DriveReady Error } Dec 1 04:40:04 Tower kernel: ata2: error=0x04 { DriveStatusError } Dec 1 04:40:04 Tower kernel: ata2: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 Dec 1 04:40:04 Tower kernel: ata2: status=0x41 { DriveReady Error } Dec 1 04:40:04 Tower kernel: ata2: error=0x04 { DriveStatusError } Dec 1 04:40:04 Tower kernel: ata2: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 Dec 1 04:40:04 Tower kernel: ata2: status=0x41 { DriveReady Error } Dec 1 04:40:04 Tower kernel: ata2: error=0x04 { DriveStatusError } Dec 1 04:40:04 Tower kernel: ata2: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 Dec 1 04:40:04 Tower kernel: ata2: status=0x41 { DriveReady Error } Dec 1 04:40:04 Tower kernel: ata2: error=0x04 { DriveStatusError } Dec 1 04:40:04 Tower kernel: ata2: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 Dec 1 04:40:04 Tower kernel: ata2: status=0x41 { DriveReady Error } Dec 1 04:40:04 Tower kernel: ata2: error=0x04 { DriveStatusError } Dec 1 04:40:04 Tower kernel: ata2: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 Dec 1 04:40:04 Tower kernel: ata2: status=0x41 { DriveReady Error } Dec 1 04:40:04 Tower kernel: ata2: error=0x04 { DriveStatusError } Dec 1 04:40:04 Tower kernel: ata2: translated ATA stat/err 0x41/04 to SCSI SK/ASC/ASCQ 0xb/00/00 Dec 1 04:40:04 Tower kernel: ata2: status=0x41 { DriveReady Error } Dec 1 04:40:04 Tower kernel: ata2: error=0x04 { DriveStatusError } Here is a link to my complete syslog: https://docs.google.com/open?id=0B0PxQbEvxyY1T0xVbC1rN3liQ0U I am running version 4.7 I tried to run a smart test (short) on this disk and got this message: smartctl -t short -d ata /dev/sdc 2>&1 smartctl 5.39.1 2010-01-28 r3054 [i486-slackware-linux-gnu] (local build) Copyright © 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net Smartctl: Device Read Identity Failed (not an ATA/ATAPI device) A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options. It appears that when checking parity, there was a failure to write to the parity drive & it was disabled. I found these corrective actions, and I would like some confirmation/advice before I proceed: >>The reason for the "write" failure could be anything from a loose cable/bad cable (power or data) or a power supply unable to keep up to the total demands of the server, to a bad sector on the disk, or a bad disk. You need to try to get a smart report from the disk. Then, stop the array, power down, re-seat the cables, and see if it subsequently responds. Then, you'll need to re-construct the disabled drive (assuming it is good)
December 2, 201213 yr Yes. Power down. Check cabling. Reboot and try to get a SMART report. Post the report here.
December 2, 201213 yr Author Thanks for the input - I did power down the array & checked for loose connections. I did not find any, but check/pushed in the applicable plugs. I power up the array & received this message: (UnRAID Main) Stopped. New parity disk installed. (MyMain) RECON_DISK, unRAID ARRAY is STOPPED 7 disks in array. PARITY NOT VALID: DISK_DSBL_NEW The parity disk is flagged as DSBL_NEW I ran a smart test on the parity disk (see attached) smartctl -a -d ata /dev/sdc | todos >/boot/smart.txt Looks like it passed - "Reallocated_Sector_Ct", the RAW_VALUE = 0 - "Current_Pending_Sector", the RAW_VALUE = 21 - "Temperature_Celsius" = 40 Any feedback on these results? It appears the 'current pending sector' value is not good news. My temperatures have been high, but primarily with the WD drives. Anything I should do? Run smart report monthly? Assuming the parity drive is Ok to return to service, what are my next steps? Press 'refresh' via UnRAID Main menu (the Start is disabled)? smart.txt
December 2, 201213 yr You must check the box under 'start' to enable it. The sectors pending reallocation on the smart report should go to zero after the reconstruction is complete.
December 2, 201213 yr Author Thanks - I pressed start & now rebuilding parity. I will report back when done !
December 3, 201213 yr Author I am back in business - parity is valid ! Thank you for your assistance !
December 3, 201213 yr I am back in business - parity is valid ! Thank you for your assistance ! yes, but have you now performed a parity "check" to ensure you can read what was just written. If you have not, you are not through... Also, AFTER the parity check, you should get another SMART report and see that no additional sectors are pending re-allocation. If there are, that disk is probably overdue for replacement. Joe L.
December 4, 201213 yr 40 deg C for a green WD in a non active state is high IMHO. Check out your cooling.
December 4, 201213 yr Author yes, but have you now performed a parity "check" to ensure you can read what was just written. If you have not, you are not through... Also, AFTER the parity check, you should get another SMART report and see that no additional sectors are pending re-allocation. If there are, that disk is probably overdue for replacement. Joe L. Thanks - good advice, and I still have problems ! You are correct - I did just rebuild parity, and did not perform a check. I now have ran a parity check and it failed. I got the same error message PARITY NOT VALID: DISK_DSBL, and my parity disk has been disabled. It seemed to be going well. I was able to refresh my screen periodically and saw that it was progressing without errors. Last I saw it was at 94%. Here is a link to an updated syslog https://docs.google.com/open?id=0B0PxQbEvxyY1V210dUdnZWZFaWc I powered down the array & restarted. I ran a smart test on the parity disk (see attached) smartctl -a -d ata /dev/sdc | todos >/boot/smart.txt The drive passed again, this time with no pending sectors - "Reallocated_Sector_Ct", the RAW_VALUE = 0 - "Current_Pending_Sector", the RAW_VALUE = 0 - "Temperature_Celsius" = 41 The temperatures have always been high with my WD drives - higher than my Seagates & Hitachi drives. The parity drive is a 2TB WD EARS-00M. I did notices the WD drives (cache, parity, storage) at ~50C during the parity check. I have wanted to review my cooliing, to make sure I have the fans correct for an efficient flow. Not sure if this is cause of my problems. Any thoughts on next steps? Try to rebuild parity again, then check again? smart2.txt
December 5, 201213 yr Author Will do - but not sure how.... I have an open slot on my NORCO Hot Swap Rack. Could I just move the drive to one of the open bays? This would (kinda of) get me a new cable. Would UnRaid still recognize this as my parity drive?
December 5, 201213 yr The drive passed again, this time with no pending sectors - "Reallocated_Sector_Ct", the RAW_VALUE = 0 - "Current_Pending_Sector", the RAW_VALUE = 0 - "Temperature_Celsius" = 41 The temperatures have always been high with my WD drives - higher than my Seagates & Hitachi drives. The parity drive is a 2TB WD EARS-00M. I did notices the WD drives (cache, parity, storage) at ~50C during the parity check. I have wanted to review my cooliing, to make sure I have the fans correct for an efficient flow. Not sure if this is cause of my problems. Any thoughts on next steps? Try to rebuild parity again, then check again? You should very seriously look at your cooling if your WD Green drives are used at 50 deg C According to this article: http://www.tomshardware.com/reviews/hdd-reliability-storelab,2681-2.html they do not like temperature more than 40 - 45 deg C You may have cooked them already.
December 5, 201213 yr Author HA! my ace detective work has developed clue to my cooling concern. The fan is not working on my Norco unit.... The fans are working on the other Norco hot swap modules, but not this one. This unit is filled with my four WD drives. I am troubleshooting this now. I have switched some cables - no luck. I am going to try a replacement fan. Temps have always been higher for my WD drives. Now I know why... Rarely at 50C, but I did notice this temp after heavy usage (rebuit parity, then check parity). This usage can take >3 hrs. Once I get my cooling fixed, I will give parity another try.
December 16, 201213 yr Author Think I got this solved! I have rebuilt & checked parity = OK. I ended up replacing the hard drive. I tried working with the old one, but could not get parity to check. I switched cables, replaced a cooling fan, swapped out Norco units, no luck. Could not get the parity to check after a rebuild. Decided to replace the hard drive when I was running out of troubleshooting steps. The old one did not fail a smart check, but could not pass parity. I got a new drive, ran preclear twice, and parity is up & fixed. Thanks for all the help !
Archived
This topic is now archived and is closed to further replies.