March 27, 201016 yr Flash share problem.. am I the only one who have problem to acces the flash share in write mode. The share is exported in read/write and everytime I try to copy, create or modify I'm able to do it like one time after I have access denied on eveymod I try to make until I reboot the server after the reboot I'll able to do 1-2 operation then access denied again till another reboot. Really anoying Odds are very high your file-system is corrupted and needs repair, or the flash drive is starting to fail. Post a syslog, it will provide the clues needed for analysis. Yes, it is just you. The only time I've ever had the same situation here is if my flash drive was full, or too many files in the root folder (after all, it is a FAT file-system) You can probably put the flash drive in your windows PC and run checkdisk or scandisk on it to fix it. The linux kernel will make the drive read-only if it detects corruption to prevent you from writing to it and causing even more corruption.
March 27, 201016 yr damn I think you're right Mar 26 19:46:52 NAS kernel: FAT: Filesystem error (dev sdd1) Mar 26 19:46:52 NAS kernel: invalid access to FAT (entry 0x83e57e1d) Mar 26 19:46:52 NAS kernel: FAT: Filesystem error (dev sdd1) Mar 26 19:46:52 NAS kernel: invalid access to FAT (entry 0x83e57e1d) Mar 26 19:46:52 NAS kernel: FAT: Filesystem error (dev sdd1) Mar 26 19:46:52 NAS kernel: invalid access to FAT (entry 0x7f4b047c) this usb key is an almost brand new kingston... and when you on my laptop in windows xp no problem to acces it and write to it
March 27, 201016 yr damn I think you're right Mar 26 19:46:52 NAS kernel: FAT: Filesystem error (dev sdd1) Mar 26 19:46:52 NAS kernel: invalid access to FAT (entry 0x83e57e1d) Mar 26 19:46:52 NAS kernel: FAT: Filesystem error (dev sdd1) Mar 26 19:46:52 NAS kernel: invalid access to FAT (entry 0x83e57e1d) Mar 26 19:46:52 NAS kernel: FAT: Filesystem error (dev sdd1) Mar 26 19:46:52 NAS kernel: invalid access to FAT (entry 0x7f4b047c) this usb key is an almost brand new kingston... and when you on my laptop in windows xp no problem to acces it and write to it Run scandisk on it. Make sure you "safely eject" it.
March 28, 201016 yr Flash share problem.. am I the only one who have problem to acces the flash share in write mode. The share is exported in read/write and everytime I try to copy, create or modify I'm able to do it like one time after I have access denied on eveymod I try to make until I reboot the server after the reboot I'll able to do 1-2 operation then access denied again till another reboot. Really anoying I had this issue as well. I think it was just a because the flash drive was substandard. I solved it by getting another one, but leaving the old one in for the license key. i modified the go script to unmount the old one (licensed faulty one), and then mount the new good one in it's place. So unraid boots off the new one, then mounts it, then the go script unmounts it, mounts the new one as boot, and continues running a go script on it. This works fine and is seemless. there is a post somewhere describing the details of it. search by my user id.
March 28, 201016 yr Flash share problem.. am I the only one who have problem to acces the flash share in write mode. The share is exported in read/write and everytime I try to copy, create or modify I'm able to do it like one time after I have access denied on eveymod I try to make until I reboot the server after the reboot I'll able to do 1-2 operation then access denied again till another reboot. Really anoying I had this issue as well. I think it was just a because the flash drive was substandard. I solved it by getting another one, but leaving the old one in for the license key. i modified the go script to unmount the old one (licensed faulty one), and then mount the new good one in it's place. So unraid boots off the new one, then mounts it, then the go script unmounts it, mounts the new one as boot, and continues running a go script on it. This works fine and is seemless. there is a post somewhere describing the details of it. search by my user id.
April 2, 201016 yr I am having the split issue as well. I also am seeing the return of the disappearing user shares. Stop & Start array and the files return.
April 5, 201016 yr I'd roll back or downgrade to 4.5.1, much better as it seems to be less bug prone. I am having the split issue as well. I also am seeing the return of the disappearing user shares. Stop & Start array and the files return.
April 8, 201016 yr I have had the disappearing files twice since I upgraded to 4.5.3 It scared the crap out of me the first time :'( I am rolling back to 4.5.1 as we speak. B2
April 8, 201016 yr Did 4.5.1 fix the problem? I'd test by restarting your server several time at random to properly test, but I'd be confident it would fix the issue as I've heard little bad press about 4.5.1. I have had the disappearing files twice since I upgraded to 4.5.3 It scared the crap out of me the first time :'( I am rolling back to 4.5.1 as we speak. B2
April 24, 201016 yr Yes reverting to 4.5.1. fixed my disappearing file problem. Was 4.5.3 suppose to address the spiky bandwidth problem? B2 Did 4.5.1 fix the problem? I'd test by restarting your server several time at random to properly test, but I'd be confident it would fix the issue as I've heard little bad press about 4.5.1. I have had the disappearing files twice since I upgraded to 4.5.3 It scared the crap out of me the first time :'( I am rolling back to 4.5.1 as we speak. B2
April 24, 201016 yr Basically the link below explains the exact reason why: http://lime-technology.com/forum/index.php?topic=5496.0 It was a minor update to support another addon card and a Kernel update too. Yes reverting to 4.5.1. fixed my disappearing file problem. Was 4.5.3 suppose to address the spiky bandwidth problem? B2 Did 4.5.1 fix the problem? I'd test by restarting your server several time at random to properly test, but I'd be confident it would fix the issue as I've heard little bad press about 4.5.1. I have had the disappearing files twice since I upgraded to 4.5.3 It scared the crap out of me the first time :'( I am rolling back to 4.5.1 as we speak. B2
April 24, 201016 yr Hi, Not sure if this is the right place, so please let me know where the right place to ask this is if I'm at the wrong place. I had installed a new 1.5TB disk which worked for a couple of weeks. Then on the last reboot of the server, the disk is mounted as read-only and I don't know what to do to fix it. I was using 4.5.3 and downgraded to 4.5.1 just to check. Attached is the syslog. Can someone please advise? Thanks! Vincent syslog.txt
April 24, 201016 yr Hi, Not sure if this is the right place, so please let me know where the right place to ask this is if I'm at the wrong place. I had installed a new 1.5TB disk which worked for a couple of weeks. Then on the last reboot of the server, the disk is mounted as read-only and I don't know what to do to fix it. I was using 4.5.3 and downgraded to 4.5.1 just to check. Attached is the syslog. Can someone please advise? Thanks! Vincent Your file system on disk7 ( /dev/md7 ) has some corruption. To repair it you will need to follow the steps shown in the wiki here: http://lime-technology.com/wiki/index.php?title=Check_Disk_Filesystems The file-system was mounted as read-only to prevent further corruption if written to in its current state. Joe L.
April 24, 201016 yr I spoke to soon. My files disappeared again and I am running 4.5.1. Yes reverting to 4.5.1. fixed my disappearing file problem. Was 4.5.3 suppose to address the spiky bandwidth problem? B2 Did 4.5.1 fix the problem? I'd test by restarting your server several time at random to properly test, but I'd be confident it would fix the issue as I've heard little bad press about 4.5.1. I have had the disappearing files twice since I upgraded to 4.5.3 It scared the crap out of me the first time :'( I am rolling back to 4.5.1 as we speak. B2
April 24, 201016 yr bburns, if I recall correctly, the syslog you posted showed a ton of DUPLICATE DIRECTORY or DUPLICATE FILE warrnings. You need to fix those first before you go on about files disappearing. I suspect that is your problem.
April 26, 201016 yr I Have a disk showing as missing after the 4.5.3 upgrade...(from 4.5) Smells alot like the old disk missing issue, but doesn't seem to show the same way... My syslog shows it's detecting the drive with a slightly different name, and the menu screen shows the disk as "Missing". Nowhere does the different name show up in the user interface (not under devices either)...the drive shows with a red bullet. The drive is a Hitachi, and I have 2 other Hitachi's in my setup and they're detected just fine...It almost looks like the newly detected name is corrupted somehow. Downgrading to 4.5 shows everything alright. I'm trying to upgrade to 4.5.3 in order to support the new supermicro board I'm about to install. Here's what my main screen shows: Version 4.5: Model / Serial No. Temperature Size Free Reads Writes Errors parity ata-Hitachi_HDS721010KLA330_GTJ100PAG0YG3C 42°C 976,762,552 - 105 122 0 disk1 ata-SAMSUNG_HD103UJ_S13PJDWS323320 34°C 976,762,552 36,868,312 54 6 0 disk2 ata-SAMSUNG_HD103UJ_S13PJ1BQ702829 33°C 976,762,552 15,414,156 55 6 0 disk3 ata-SAMSUNG_HD103UJ_S13PJDWS323304 30°C 976,762,552 38,589,180 52 6 0 disk4 ata-SAMSUNG_HD103UJ_S13PJ1BQ735050 34°C 976,762,552 140,416,300 52 6 0 disk5 ata-Hitachi_HDS721010KLA330_GTG000PAG17Z4C 39°C 976,762,552 85,548,068 53 6 0 disk6 ata-Hitachi_HDS721010KLA330_GTJ100PAG0HBSC 40°C 976,762,552 636,785,268 118 78 0 disk7 ata-Hitachi_HDT721010SLA360_STF604MR2MDJ6P 35°C 976,762,552 422,973,780 48 5 0 disk8 ata-Hitachi_HDT721010SLA360_STF604MR3GM8SP 34°C 976,762,552 937,514,212 52 7 0 disk9 Not installed disk10 Not installed ----------------------------------------- Post 4.5.3 upgrade Model / Serial No. Temperature Size Free Reads Writes Errors parity Hitachi_HDS72101_GTJ100PAG0YG3C 42°C 976,762,552 - - - - disk1 SAMSUNG_HD103UJ_S13PJDWS323320 34°C 976,762,552 - - - - disk2 SAMSUNG_HD103UJ_S13PJ1BQ702829 33°C 976,762,552 - - - - disk3 SAMSUNG_HD103UJ_S13PJDWS323304 30°C 976,762,552 - - - - disk4 SAMSUNG_HD103UJ_S13PJ1BQ735050 34°C 976,762,552 - - - - disk5 Hitachi_HDS72101_GTG000PAG17Z4C 38°C 976,762,552 - - - - disk6 Hitachi_HDS72101_GTJ100PAG0HBSC 40°C 976,762,552 - - - - disk7 Missing Hitachi_HDT721010SLA360_STF604MR2MDJ6P - - 976,762,552 - - - - disk8 Hitachi_HDT72101_STF604MR3GM8SP 34°C 976,762,552 - - - - disk9 Not installed disk10 Not installed I'm also attaching my syslogs... ...I think the entry of note in the 4.5.3 log is as follows: Apr 26 13:21:18 Tower kernel: ata1.00: model number mismatch 'Hdtachi HQT721010^BLA360 0' != 'Hitachi HDT721010SLA360' Not sure why this drive is being detected as a Hdtachi instead of Hitachi under 4.5.3 Any ideas? David syslog.45.zip syslog.453.zip
April 27, 201016 yr I did not post a syslog file nor do I recall seeing DUPLICATE_* warnings at the time. But thanks for the tip on what to look for. I went and looked at my previous syslog and did see shfs: dunplicate object: messages. So how do dupe files occur? I assume it does not matter which copy is deleted? Also I am pretty sure that the files that are duped existed prior to my upgrade to 4.5.1. What is the best way to find all the duplicate files? bburns, if I recall correctly, the syslog you posted showed a ton of DUPLICATE DIRECTORY or DUPLICATE FILE warnings. You need to fix those first before you go on about files disappearing. I suspect that is your problem.
April 27, 201016 yr d = 100 = 1100100 i = 105 = 1101001 Q = 81 = 1010001 D = 68 = 1000100 B = 66 = 1000010 S = 83 = 1010011 Not an obvious memory coruption issue, I'm assuming it is repeatable from your description? Does it always identify the drive as Hdtachi HQT721010^BLA360? From the console can you try a hdparm -i /dev/sd? where ? = drive letter for drive 7 (unRAID devices page in 4.5 will show this info). It may well be slackware rather than unRAID issue. Also what controller is the drive attached too?
April 29, 201016 yr Well to me this is interesting, effect of the latest updates on my hardware when doing a parity check I'm getting these numbers 4.5.3 18 KB/sec 4.5.1 26,000 KB/sec 4.2.2 38,000 KB/sec Sam
April 30, 201016 yr Kaygee: The drive always reports the same way every reboot... It doesn't get assigned a device identifyer, so I'm not sure I can use the hdparm command. When I use hdparm under 4.5 it reports correctly, but under 4.5.3, a different drive is assigned it's drive letter, and it doesn't get one... The drive is connected to a Sil controller (two port, pci-e controller). This controller also has another Hitachi drive connected to it, which is working correctly. d = 100 = 1100100 i = 105 = 1101001 Q = 81 = 1010001 D = 68 = 1000100 B = 66 = 1000010 S = 83 = 1010011 Not an obvious memory coruption issue, I'm assuming it is repeatable from your description? Does it always identify the drive as Hdtachi HQT721010^BLA360? From the console can you try a hdparm -i /dev/sd? where ? = drive letter for drive 7 (unRAID devices page in 4.5 will show this info). It may well be slackware rather than unRAID issue. Also what controller is the drive attached too?
May 1, 201016 yr I moved the drive in question to another controller and it's reading fine... not sure why one channel of this sil controller seems to have a problem with 4.5.3 David
May 1, 201016 yr Here is an update. I removed all duplicate files (the ones found by Unmenu's dupe file page and what was being noted in the logfile). I am still seeing the no file problem. Here is the log file from the last time it happened. Note that prior to the point that I snipped the log there were many many messages about "Too many open files". When I stopped the array, Disk 1 would not unmount. Required a power cycle to recover. Apr 30 08:17:50 Ungol shfs: shfs_open: open: /mnt/disk8/Pictures/2008December/Hawaii/febo9 286.JPG (24) Too many open files Apr 30 08:17:52 Ungol in.telnetd[9282]: connect from 192.168.15.205 (192.168.15.205) Apr 30 08:17:58 Ungol shfs: shfs_open: open: /mnt/disk8/Pictures/2001May/DisneyWithBiel/231-3182_IMG.JPG (24) Too many open files Apr 30 08:18:02 Ungol shfs: shfs_open: open: /mnt/disk8/Pictures/2001May/DisneyWithBiel/231-3182_IMG.JPG (24) Too many open files Apr 30 08:18:04 Ungol login[9283]: ROOT LOGIN on `pts/1' from `192.168.15.205' . . Apr 30 08:23:45 Ungol emhttp: Retry unmounting disk share(s)... Apr 30 08:23:50 Ungol emhttp: shcmd (190): umount /mnt/disk1 >/dev/null 2>&1 Apr 30 08:23:50 Ungol emhttp: _shcmd: shcmd (190): exit status: 1 Apr 30 08:23:50 Ungol emhttp: shcmd (191): rmdir /mnt/disk1 >/dev/null 2>&1 Apr 30 08:23:50 Ungol emhttp: _shcmd: shcmd (191): exit status: 1 Apr 30 08:23:50 Ungol emhttp: Retry unmounting disk share(s)... Apr 30 08:23:55 Ungol emhttp: shcmd (192): umount /mnt/disk1 >/dev/null 2>&1 Apr 30 08:23:55 Ungol emhttp: _shcmd: shcmd (192): exit status: 1 Apr 30 08:23:55 Ungol emhttp: shcmd (193): rmdir /mnt/disk1 >/dev/null 2>&1 Apr 30 08:23:55 Ungol emhttp: _shcmd: shcmd (193): exit status: 1 Apr 30 08:23:55 Ungol emhttp: Retry unmounting disk share(s)... Apr 30 08:24:00 Ungol emhttp: shcmd (194): umount /mnt/disk1 >/dev/null 2>&1 Apr 30 08:24:00 Ungol emhttp: _shcmd: shcmd (194): exit status: 1 Apr 30 08:24:00 Ungol emhttp: shcmd (195): rmdir /mnt/disk1 >/dev/null 2>&1 Apr 30 08:24:00 Ungol emhttp: _shcmd: shcmd (195): exit status: 1 Apr 30 08:24:00 Ungol emhttp: Retry unmounting disk share(s)... I did not post a syslog file nor do I recall seeing DUPLICATE_* warnings at the time. But thanks for the tip on what to look for. I went and looked at my previous syslog and did see shfs: dunplicate object: messages. So how do dupe files occur? I assume it does not matter which copy is deleted? Also I am pretty sure that the files that are duped existed prior to my upgrade to 4.5.1. What is the best way to find all the duplicate files? bburns, if I recall correctly, the syslog you posted showed a ton of DUPLICATE DIRECTORY or DUPLICATE FILE warnings. You need to fix those first before you go on about files disappearing. I suspect that is your problem.
May 1, 201016 yr Here is an update. I removed all duplicate files (the ones found by Unmenu's dupe file page and what was being noted in the logfile). I am still seeing the no file problem. Here is the log file from the last time it happened. Note that prior to the point that I snipped the log there were many many messages about "Too many open files". When I stopped the array, Disk 1 would not unmount. Required a power cycle to recover. Apr 30 08:17:50 Ungol shfs: shfs_open: open: /mnt/disk8/Pictures/2008December/Hawaii/febo9 286.JPG (24) Too many open files Apr 30 08:17:52 Ungol in.telnetd[9282]: connect from 192.168.15.205 (192.168.15.205) Apr 30 08:17:58 Ungol shfs: shfs_open: open: /mnt/disk8/Pictures/2001May/DisneyWithBiel/231-3182_IMG.JPG (24) Too many open files Apr 30 08:18:02 Ungol shfs: shfs_open: open: /mnt/disk8/Pictures/2001May/DisneyWithBiel/231-3182_IMG.JPG (24) Too many open files Apr 30 08:18:04 Ungol login[9283]: ROOT LOGIN on `pts/1' from `192.168.15.205' . . Apr 30 08:23:45 Ungol emhttp: Retry unmounting disk share(s)... Apr 30 08:23:50 Ungol emhttp: shcmd (190): umount /mnt/disk1 >/dev/null 2>&1 Apr 30 08:23:50 Ungol emhttp: _shcmd: shcmd (190): exit status: 1 Apr 30 08:23:50 Ungol emhttp: shcmd (191): rmdir /mnt/disk1 >/dev/null 2>&1 Apr 30 08:23:50 Ungol emhttp: _shcmd: shcmd (191): exit status: 1 Apr 30 08:23:50 Ungol emhttp: Retry unmounting disk share(s)... Apr 30 08:23:55 Ungol emhttp: shcmd (192): umount /mnt/disk1 >/dev/null 2>&1 Apr 30 08:23:55 Ungol emhttp: _shcmd: shcmd (192): exit status: 1 Apr 30 08:23:55 Ungol emhttp: shcmd (193): rmdir /mnt/disk1 >/dev/null 2>&1 Apr 30 08:23:55 Ungol emhttp: _shcmd: shcmd (193): exit status: 1 Apr 30 08:23:55 Ungol emhttp: Retry unmounting disk share(s)... Apr 30 08:24:00 Ungol emhttp: shcmd (194): umount /mnt/disk1 >/dev/null 2>&1 Apr 30 08:24:00 Ungol emhttp: _shcmd: shcmd (194): exit status: 1 Apr 30 08:24:00 Ungol emhttp: shcmd (195): rmdir /mnt/disk1 >/dev/null 2>&1 Apr 30 08:24:00 Ungol emhttp: _shcmd: shcmd (195): exit status: 1 Apr 30 08:24:00 Ungol emhttp: Retry unmounting disk share(s)... I did not post a syslog file nor do I recall seeing DUPLICATE_* warnings at the time. But thanks for the tip on what to look for. I went and looked at my previous syslog and did see shfs: dunplicate object: messages. So how do dupe files occur? I assume it does not matter which copy is deleted? Also I am pretty sure that the files that are duped existed prior to my upgrade to 4.5.1. What is the best way to find all the duplicate files? bburns, if I recall correctly, the syslog you posted showed a ton of DUPLICATE DIRECTORY or DUPLICATE FILE warnings. You need to fix those first before you go on about files disappearing. I suspect that is your problem. Typically, if a disk will not un-mount it is currently "busy" It could be busy because it is the "current directory" for a process. (You had "cd'd" there before starting a program that is still running) Or you might have logged in and changed directory to it ("cd /mnt/disk1" will make disk1 busy since it is your current directory) A disk can also be busy if a file on it is open for reading or writing. (If the mover is moving a file onto it from the cache drive it will be busy) A disk can also be busy if a directory on it has been used as a mount point for another file-system. See this post for more detail: http://lime-technology.com/forum/index.php?topic=4633.msg60316#msg60316 Joe L.
May 1, 201016 yr @David If it was a driver issue I would of thought we would of heard of it by now. A lot of people use SIL 3132 based cards. Are the drive firmwares the same? Could be a bad install, try re-download. Try the other hitachi on that channel or another drive. Do a lot of testing before trusting that channel though. Data corruption can be incidious. Copy, move, create and delete and md5 each data set after each operation. I was suprised at your previous answer that the drive was seen by hdparm this hints that slackware isnt mounting the device. From a console prompt ls /dev/sd? under 4.5 and 4.5.3 should list all sata devices.
Archived
This topic is now archived and is closed to further replies.