October 9, 20169 yr Passed the new update last night and found that all my shared were missing. I can access all the drives by themselves just find but it wont export my shares. I have tried to change the global share settings but still no shares. This causes docker and virtual machines to not function and will make the server hang when they are attempted to be started since they use a share to access there files. I was forced to downgrade back to 6.1.9 and recreate my docker apps to make things work again. Please advise how to proceed. I have attached a diagnostic file for reference. Thanks M/B: MSI - Z97 GAMING 5 (MS-7917) CPU: Intel core i7-4790k CPU @ 4.00Ghz MEMORY: 32768 MB Network: bond0 IEEE 802.3ad Dynamic Link Aggregation Disks: 9 Data--> 2TB western Digital Red Pro 1 Parity--> 2TB western digital Red Pro 2 Cache --> 250GB Crucial SSD tower-diagnostics-20161008-2106.zip
October 9, 20169 yr Everything looks perfect in your logs except there are no shares Please boot back into 6.2.1, verify still no shares, then open ssh/telnet session and post the output of these commands: v /mnt/user v /mnt/user/* Those two commands will list the names of your user shares. Instead of posting that here you can email to me: [email protected]
October 9, 20169 yr Author I sent them to your email. It appears that everything is there, but nothing is showing in the GUI and nothing is exported. If i try to create a share, it says share deleted but it shows when i list the shares in terminal. Never seen this before.
October 9, 20169 yr Author not sure what to do next, i put the system in safe mode and still no shares being exported. Do I need to try for a clean install?
October 9, 20169 yr Author all my dockers run from the cache pool drives, if i unassigned them, i wont be able to run any of those. I will try to unassign them to see what happens.
October 10, 20169 yr Author okay, so if I unassign my cache pool drives i get my shares back. However, all my applications run from the cache pool so how do i fix this?
October 10, 20169 yr Apparently you have some flavor of corruption on your cache pool. You probably should back up everything on the cache pool to another drive, and then run the BTRFS file system checks, starting with a scrub. Worst case scenario, reformat and copy your backup back on. Have you had any unclean shutdowns, or power cuts not handled by your UPS? BTRFS can be fragile if not shut down properly.
October 10, 20169 yr Author I have had several unclean shutdowns due to server hanging and becoming unresponsive. I don't understand how this corruption, if that is what has happened, would cause my shares to dissapear. What is weared about this is that everything works fine in the previous version and its only when i upgrade to this new release i am having problems. I tried to run scrub with the option checked to fix file system errors and still no dice. I also tried to run balance and then remove only 1 of my disks from the pool but then the disk by itself was unreadable and wanted to be formatted. I will probably try to move the data off the disk tonight and format them to see what happens. If this doesn't work, I'm not sure what else can be done. Thanks.
October 10, 20169 yr I also tried to run balance and then remove only 1 of my disks from the pool but then the disk by itself was unreadable and wanted to be formatted. Who have to physically disconnect the disk, simply unassigning it will result in an unmountable cache.
October 11, 20169 yr Author well, just disconnecting one or the other disks gave me the same result; No Shares. I copied everything off the cache pool and reformatted the cache disks one at a time and then verified that my shares were still available, which they were. I then copied everything back to the cache pool and setup my dockers and vm's. Everything works now except for one of my VM's. When I try to start it, it locks up the server and forces me to do a Forced Restart. If I try to force the vm to stop i get an error about it not being able to terminate the process. Any Ideas? I still cant wrap my head around why the cache disks would keep my shares from showing up. Everything could be accessed from the cache pool just fine Can someone give me some insight on this?
October 24, 20169 yr Hi all, I just tried to update from 6.1.9 to 6.2.1 and have run into the exact same issue, however I am not using a cache pool, just a single cache drive. I've tried scrubbing the cache drive, no errors were found. I've tried unassigning the cache drive, still no user shares. All data is present on the array (verified via ssh, and that my Dockers that reference the user shares locally work), but none of the user shares are exposed in the web UI or via Samba/CIFS. I've attached the diagnostics, any help would be appreciated. unraid-diagnostics-20161023-2238.zip
October 25, 20169 yr I got a chance to try the formatting the cache drive trick. For no apparent reason, formatting the cache drive made things "better", the shares now show up in the web UI, but they are still not accessible over SAMBA. -edit: scratch that, they're gone again from the web UI. I'm kind of at a loss here. Guess I'll roll back to 6.1.9 unless anyone has any ideas. unraid-diagnostics-20161024-2003.zip
October 25, 20169 yr I got a chance to try the formatting the cache drive trick. For no apparent reason, formatting the cache drive made things "better", the shares now show up in the web UI, but they are still not accessible over SAMBA. -edit: scratch that, they're gone again from the web UI. I'm kind of at a loss here. Guess I'll roll back to 6.1.9 unless anyone has any ideas. I took a look, but completely drawing a blank. Syslog shows no issues at all that I could see. CacheDirs is seeing top level folders, the config/shares folder is full of share settings files, User Shares are enabled, shareSMBEnabled is yes, shareSMBCount is 13, but there are no shares listed anywhere in vars and shareCount is 0. shareCount being zero is the visible part of the issue. I think bonienl or Tom or Eric will need to take a look. cybertronics appears to have the same situation. Edit: just discovered Tom is already on it!
October 25, 20169 yr Edit: just discovered Tom is already on it! Nice! Let me know if there's anything I can try, or any info I can provide. I'll try to get any requests turned around quickly, but this is my production server, running my DVR (SageTV) and some other "important" stuff, so I have to be careful when I take it down.
October 25, 20169 yr Author I had to completely wipe my cache drives before unraid would allow me to reformat them. after the reformat, everything was good except for a windows 7 VM that refused to run. VM would stop responding in the middle of the boot process and when i would try to force it to stop, the whole server would stop responding and crash.
October 26, 20169 yr I just opted to change the filesystem from BTRFS to XFS, so that allowed me to format it. Unfortunately for me, that didn't help, still no shares until I went back to 6.1.
October 26, 20169 yr I just opted to change the filesystem from BTRFS to XFS, so that allowed me to format it. Unfortunately for me, that didn't help, still no shares until I went back to 6.1. When you get a chance, reboot into 6.2.x, Start array then and post output of these two commands: v /mnt v /mnt/user Also please try booting into 'safe mode' once so we can rule out any plugin incompatibilities.
October 26, 20169 yr Here ya go: v /mnt drw-rw-rw- 7 nobody users 85 Oct 26 03:40 cache/ drw-rw-rw- 10 nobody users 151 Oct 25 22:02 disk1/ drw-rw-rw- 11 nobody users 168 Jul 16 05:25 disk10/ drw-rw-rw- 11 nobody users 181 Jul 16 06:08 disk11/ drw-rw-rw- 11 nobody users 184 Jul 15 16:55 disk2/ drw-rw-rw- 11 nobody users 4096 Aug 16 17:53 disk3/ drw-rw-rw- 13 nobody users 220 Jul 26 03:40 disk4/ drw-rw-rw- 12 nobody users 190 Jul 16 10:05 disk5/ drw-rw-rw- 13 nobody users 207 Jul 16 11:34 disk6/ drw-rw-rw- 8 nobody users 116 Nov 13 2015 disk7/ drw-rw-rw- 8 nobody users 117 Jul 4 03:40 disk8/ drw-rw-rw- 9 nobody users 134 Oct 23 19:46 disk9/ drw-rw-rw- 1 nobody users 85 Oct 26 03:40 user/ drw-rw-rw- 1 nobody users 151 Oct 25 22:02 user0/ v /mnt/user drw-rw-rw- 1 nobody users 23 Oct 23 18:18 Backup/ drwxrwxrwx 1 nobody users 203 Oct 25 17:57 Blu-Ray/ drwxrwxrwx 1 nobody users 4096 Jul 14 2015 Blu-ray_TV/ drwxrwxrwx 1 nobody users 194 Dec 8 2015 DVDs/ drwxrwxrwx 1 nobody users 4096 Apr 4 2016 HDDVD/ drwxrwxrwx 1 nobody users 99 Oct 26 03:40 Media/ drwxrwxrwx 1 nobody users 4096 Jan 31 2016 Private/ drwxrwxrwx 1 nobody users 60 Oct 16 20:34 Recordings/ drwxrwx--- 1 nobody users 211 Oct 12 21:02 TV/ drw-rw-rw- 1 nobody users 161 Oct 24 20:48 appdata/ drw-rw-rw- 1 nobody users 6 Oct 21 03:40 apps/ drwxrwxrwx 1 nobody users 87 Jul 7 15:23 cache_backup/ drwxrwxrwx 1 nobody users 87 May 21 2015 crashplan/ drw-rw-rw- 1 nobody users 20 Jul 7 16:19 crashplan.bak/ -rw-rw-rw- 1 nobody users 335872 Feb 28 2012 dupes_bd.txt -rw-rw-rw- 1 nobody users 69385542 Feb 28 2012 dupes_out.txt -rw-rw-rw- 1 nobody users 69385542 Feb 28 2012 dupes_out.txt.bak -rw-rw-rw- 1 nobody users 70286280 Feb 28 2012 dupes_out_dos.txt -rw-rw-rw- 1 nobody users 208637923 Feb 26 2012 dupes_tmp1 -rw-rw-rw- 1 nobody users 169841688 Feb 26 2012 dupes_tmp2 -rw-rw-rw- 1 nobody users 222601989 Feb 27 2012 dupes_tmp3 -rw-rw-rw- 1 nobody users 70688481 Feb 27 2012 dupes_tmp4 -rw-rw-rw- 1 nobody users 91015114 Feb 28 2012 dupes_tmp5 -rw-rw-rw- 1 root root 519 Oct 23 2015 result.log drw-rw-rw- 1 nobody users 67 Oct 24 21:38 stuff/ drwxrwxrwx 1 nobody users 21 Sep 27 18:37 system/ drwxrwxrwx 1 nobody users 60 Oct 16 2010 thunderbird-backup/ drw-rw-rw- 1 nobody users 48 Jul 7 16:27 vms/ I tried safe mode, no change. I even tried deleting those odd text files (I think they were from a dup finder script), but that didn't help.
October 26, 20169 yr Here ya go: Thanks, that all looks ok. What happens if you go to Shares and try to create a new share? For example, first click on the 'Log' button on the right of the menu bar and then go to Shares/Add Share. Name it anything, maybe 'test', then click Add Share. Does the share show up? Please select/copy/paste what got generated in the log and post here.
October 26, 20169 yr So, here's what I got in the log creating the share: Oct 26 18:26:28 unRAID emhttp: shcmd (187): mkdir '/mnt/user/test' |& logger Oct 26 18:26:28 unRAID emhttp: shcmd (188): chmod 0777 '/mnt/user/test' Oct 26 18:26:28 unRAID emhttp: shcmd (189): chown 'nobody':'users' '/mnt/user/test' However the very odd thing is that after making the share, the Share settings page says that the share's been deleted (see attached picture). I get the same message if I change the URL to another share that should exist.
October 26, 20169 yr Here ya go: Wait! Hold the phone! Big glaring problem here: drw-rw-rw- 1 nobody users 85 Oct 26 03:40 user/ That should read: drwxrwxrwx 1 nobody users 85 Oct 26 03:40 user/ When unRAID processes user shares, it verifies 'x' bit is set so that files can be created in it. If the 'owner' execute permission bit is not set, it does not treat the directory as a share. If you do this: chmod 777 /mnt/user It will magically solve the mystery of the missing shares (might need to Stop/Start array). However, I notice underneath /mnt/user there are some more directories there without 'x' bit set - those will still be missing. Question is, how did those permissions get changed? Because unRAID always sets '777' for directories and '666' for files it creates. Something or someone had to explicitly change those bits.... ring a bell?
October 26, 20169 yr Wait! Hold the phone! Big glaring problem here: drw-rw-rw- 1 nobody users 85 Oct 26 03:40 user/ That should read: drwxrwxrwx 1 nobody users 85 Oct 26 03:40 user/ When unRAID processes user shares, it verifies 'x' bit is set so that files can be created in it. If the 'owner' execute permission bit is not set, it does not treat the directory as a share. If you do this: chmod 777 /mnt/user It will magically solve the mystery of the missing shares (might need to Stop/Start array). However, I notice underneath /mnt/user there are some more directories there without 'x' bit set - those will still be missing. That was easy. Thanks for the help! I chmod'd /mnt/usr and all the directories in it, and now all looks well. Question is, how did those permissions get changed? Because unRAID always sets '777' for directories and '666' for files it creates. Something or someone had to explicitly change those bits.... ring a bell? Only very, very faint bells. The only things I can think of are, a very, very long while ago I had some trouble with an unclean shutdown, no idea of specifics, I have vague recollections of having permissions problems with a docker as well. Perhaps more questionable is when I converted to XFS, it's possible I clobbered something that time. I really don't know. I know I didn't do it on purpose. Anyway, looks like I'm up and running again, and up to 6.2.2 now. Thanks again for the help!
October 27, 20169 yr Looks like this would be another check for Squid to add to his Fix Common Problems plugin
October 27, 20169 yr Looks like this would be another check for Squid to add to his Fix Common Problems plugin Maybe but would be better to find source of problem rather than mask it.
Archived
This topic is now archived and is closed to further replies.