fonzie Posted January 30, 2013 Share Posted January 30, 2013 I hadn't run a parity check in quite a while. I ran it two days ago and when it was done, it said I had 3 errors. How do I find what those errors were and if they are serious? Does unRAID automatically correct those errors, or will they show up again the next time I run a parity check? I've never gotten parity errors, so I'm not sure what to do. my syslog says: Parity updated 3 times to address sync errors. Jan 30 04:40:01 media syslogd 1.4.1: restart. Jan 30 11:44:54 media kernel: mdcmd (469): spindown 0 (Routine) Jan 30 11:44:54 media kernel: mdcmd (470): spindown 7 (Routine) Jan 30 11:44:55 media kernel: mdcmd (471): spindown 8 (Routine) Jan 30 11:45:06 media emhttp: shcmd (261): /usr/sbin/hdparm -y /dev/sdk >/dev/null (Drive related) Jan 30 11:53:35 media unmenu[1550]: which: no bwm-ng in (/bin:/usr/bin:/sbin:/usr/sbin) Link to comment
dgaschk Posted January 30, 2013 Share Posted January 30, 2013 Attach a syslog. Check all SMART reports for current_pending_sector RAW_VALUE greater than 0. Link to comment
fonzie Posted January 30, 2013 Author Share Posted January 30, 2013 I downloaded the syslog through unMenu and those few lines I posted above were all that it had. I thought there would be more, but I guess not. also, how would I go about checking all SMART reports? There is a SMART.txt file in my "Flash" directory but it seems like it is a SMART test from my parity drive only and I think it is from a long time ago. Last date modified was in 9-7-2011 Link to comment
hypyke Posted January 30, 2013 Share Posted January 30, 2013 Install Smart history and run it from unmenu Link to comment
dgaschk Posted January 30, 2013 Share Posted January 30, 2013 See here to obtain a syslog: http://lime-technology.com/forum/index.php?topic=9880.0 Search the wiki for SMART. Link to comment
fonzie Posted January 31, 2013 Author Share Posted January 31, 2013 I ran Smart History and all of my disks had a current_pending_sector value of 0 on my syslog I just have a bunch of minor issues of duplicate files that show similar to this: Jan 30 17:08:55 media shfs: duplicate object: /mnt/disk1/Movies/40 Days and 40 Nights/banner.jpg (Minor Issues) Jan 30 17:08:55 media shfs: duplicate object: /mnt/disk1/Movies/40 Days and 40 Nights/banner.jpg (Minor Issues) the ones that do concern me are the 3 that show: Jan 30 15:54:15 media emhttp: shcmd (266): /usr/sbin/hdparm -y /dev/sdk >/dev/null (Drive related) Jan 30 17:35:20 media emhttp: shcmd (267): /usr/sbin/hdparm -y /dev/sdk >/dev/null (Drive related) Jan 30 18:01:20 media emhttp: shcmd (268): /usr/sbin/hdparm -y /dev/sdk >/dev/null (Drive related) sdk is my cache drive, by the way Link to comment
FreeMan Posted January 31, 2013 Share Posted January 31, 2013 I just got two of those messages tonight: Jan 30 23:38:08 NAS emhttp: shcmd (140): /usr/sbin/hdparm -y /dev/sdh >/dev/null (Drive related) Jan 30 23:38:19 NAS emhttp: shcmd (141): /usr/sbin/hdparm -y /dev/sdh >/dev/null (Drive related) The odd things is that /dev/sdh is my cache drive, and nothing's been written to it yet since I installed it yesterday. My last parity check was 2 weeks ago, should I run another now to be safe? EDIT: forgot to include the SMART report - attached. there are 0 Current_Pending_Sectors Raw value. It is interesting to me how many of the Types are listed as "Old_age". This disk is brand new. It came in a Zotac ZBOX that I setup to run OpenELEC. I run OE off an SD card, so I pulled the drive to use as my cache. Opened the computer on Christmas day. Pulled the drive a couple of days later and set it aside until I got my unRAID built. I threw it in my WinXP machine for a few days to store some files as I cleared and transferred disks to unRAID, and now it's serving cache duty. It was precleared 2x, and it sat with some data on it for a few days. "Old_age" I do not think it means what I think it means... cache_drive_SMART_report_20130130.txt Link to comment
Joe L. Posted January 31, 2013 Share Posted January 31, 2013 I ran Smart History and all of my disks had a current_pending_sector value of 0That is good on my syslog I just have a bunch of minor issues of duplicate files that show similar to this: Jan 30 17:08:55 media shfs: duplicate object: /mnt/disk1/Movies/40 Days and 40 Nights/banner.jpg (Minor Issues) Jan 30 17:08:55 media shfs: duplicate object: /mnt/disk1/Movies/40 Days and 40 Nights/banner.jpg (Minor Issues) That indicates you have that file on disk1/Movies and ALSO on another diskX/Movies You should determine which is the one you want to keep and remove/rename the other as the one on the lowest numbered disk is the one that will be used. the ones that do concern me are the 3 that show: Jan 30 15:54:15 media emhttp: shcmd (266): /usr/sbin/hdparm -y /dev/sdk >/dev/null (Drive related) Jan 30 17:35:20 media emhttp: shcmd (267): /usr/sbin/hdparm -y /dev/sdk >/dev/null (Drive related) Jan 30 18:01:20 media emhttp: shcmd (268): /usr/sbin/hdparm -y /dev/sdk >/dev/null (Drive related) Why do those concern you? Those are commands to spin down the cache drive? Do you want it to stay spinning when idle? Link to comment
Joe L. Posted January 31, 2013 Share Posted January 31, 2013 I just got two of those messages tonight: Jan 30 23:38:08 NAS emhttp: shcmd (140): /usr/sbin/hdparm -y /dev/sdh >/dev/null (Drive related) Jan 30 23:38:19 NAS emhttp: shcmd (141): /usr/sbin/hdparm -y /dev/sdh >/dev/null (Drive related) The hdparm command is used to spin down the cache drive. The odd things is that /dev/sdh is my cache drive, and nothing's been written to it yet since I installed it yesterday. Why is it odd? It indicates it was idle and could be spun down. My last parity check was 2 weeks ago, should I run another now to be safe? many of us run it once a month in NOCORRECT mode to catch issue before they become problems. EDIT: forgot to include the SMART report - attached. there are 0 Current_Pending_Sectors Raw value. It is interesting to me how many of the Types are listed as "Old_age". This disk is brand new. It came in a Zotac ZBOX that I setup to run OpenELEC. I run OE off an SD card, so I pulled the drive to use as my cache. Opened the computer on Christmas day. Pulled the drive a couple of days later and set it aside until I got my unRAID built. I threw it in my WinXP machine for a few days to store some files as I cleared and transferred disks to unRAID, and now it's serving cache duty. It was precleared 2x, and it sat with some data on it for a few days. "Old_age" I do not think it means what I think it means... Correct. There are two categories of SMART parameters. There are those that are based on measured performance and those that are based on mechanical wear. Those that are "mechanical wear" are the Old_age category. For example, I have a disk in one old PC that has been spinning for about 10 years. Its disk is failing the "Power on hours" SMART parameter. There is nothing wrong with the disk, but it has gone over the number of hours set by the manufacturer for its expected lifetime. For your level of understanding, IGNORE all the RAW values except temperature, re-allocated sectors, and sectors pending re-allocation. Then, ONLY look to see if the NORMALIZED value is at or below its affiliated failure threshold. Joe L. Link to comment
FreeMan Posted February 1, 2013 Share Posted February 1, 2013 Thanks for the explanations, Joe, most helpful. I do have the monthly parity check package installed from unMENU, so I guess that'll be kicking off within the next 24 hours or so. I'll not worry too much, then. Link to comment
fonzie Posted February 6, 2013 Author Share Posted February 6, 2013 That indicates you have that file on disk1/Movies and ALSO on another diskX/Movies You should determine which is the one you want to keep and remove/rename the other as the one on the lowest numbered disk is the one that will be used. Is there a simple way to find which disk has the duplicate file or do I have to go browsing one by one? Additionally, can I remove all duplicates with the push of a button (I'm guessing no, but hoping for the best)? Link to comment
JonathanM Posted February 6, 2013 Share Posted February 6, 2013 Is there a simple way to find which disk has the duplicate file or do I have to go browsing one by one? Additionally, can I remove all duplicates with the push of a button (I'm guessing no, but hoping for the best)? The log tells you that the file NAME is duplicated, the file content may or may not be duplicated. The easiest way to find the files is to temporarily map all the disk shares to drive letters and do a search for the name listed in the log file. Then you can open the two (or more) files and determine which version you want to keep. Link to comment
Joe L. Posted February 6, 2013 Share Posted February 6, 2013 That indicates you have that file on disk1/Movies and ALSO on another diskX/Movies You should determine which is the one you want to keep and remove/rename the other as the one on the lowest numbered disk is the one that will be used. Is there a simple way to find which disk has the duplicate file or do I have to go browsing one by one? Additionally, can I remove all duplicates with the push of a button (I'm guessing no, but hoping for the best)? Since you seem to have unMENU installed, there is a "Dupe Files" link on its menu. It will print the locations of the files. As already mentioned, the contents may differ, so examine the files individually before deciding which to delete. Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.