AndroidCat

It is possible, just create NFS share under unraid and mount it to ESXi.

You need a DNS somewhere in your network to resolve "tower" to corresponding IP address. There is usually a simple DNS in a router provided by ISP. From my experience with Docker it is not enough to use /etc/hosts on a machine hosting dockers (unraid in your case). There are some tricks to pass that /etc/hosts to docker, but easiest way is to use your own DNS either running on a router or on some server.

You need DNS to translate tower to corresponding IP address. Usually there is a simple DNS in a router you've got from your ISP.

Did anybody figure out how "NFS Export" is supposed to work? When I enable this option in "Settings" I see no way from the GUI to export a device under NFS. Also a bug I suppose: I am unable to edit remote smb share mountpoint name. Even though I edit the mount point name from GUI it always reverts to system-assigned "smb_<server>_<share>" when I click "mount" The mountpoint name however gets changed in remote_config.cfg as mountpoint="/mnt/disks/<desired name>", but that is ignored even after reboot. Device keeps mounting under system-assigned name despite specifying custom"mountpoint". So really two bugs: no ability to edit remote share name mountpoint from gui and ignoring custom mountpoint. Unraid 6.1.6 plugin v 2015.11.18 Thanks.

Updating to v.22b wiped all my samba shares. It also set one local disk sharing to OFF. Otherwise all good

You Sir are awesome.

Thank you Sir.

You have to provide hostname which can be resolved to IP address. Resolution can happen e.g in your home router. In other words Observium only accepts hostnames.

Community Applications - as already noted should become primary way to find and add Dockers, Preclear Plugin - reliable and well integrated into dynamix, Unassigned Devices - another excellent work of gfjardim and must-have, SNMP -should be baked-in to provide monitoring data. the below could also be included: PowerDown - essential with UPS, OpenVM Tools - VMWare tools essential for ESXi users

Look in plugins section. There is SNMP plugin I successfully use with Observium and Zabbix. But at the same time I think SNMP should be included into unraid core.

Splunk docker does not run under btrfs. I had to reformat mine to xfs and restore all the dockers. Just FYI if you're going to use splunk. But that's weird app problem not filesystem issue.

Great work! Very useful application. Do you know why Tonido docker ignores (or maybe any docker in general) read-only volume mapping? In my case /unRAID <-> /mnt/disks/ is mapped as :ro but I can still delete/write files there from within Tonido. I checked how the docker gets invoked and it indeed uses :ro following what had been configured in "Volume Mappings" Does it somehow ignores read-only statement because /mnt/disks contains samba mountpoints? ls -al /mnt/disks total 8 drwxrwxrwx 6 root root 120 Jul 6 19:12 ./ drwxr-xr-x 14 root root 280 Jul 6 19:12 ../ drwxrwxrwx 3 root root 4096 Dec 31 1969 BOOTUNRAID/ drwxrwxrwx 2 root root 0 Jul 6 03:40 unraid_unraid/ drwxrwxrwx 2 root root 0 Jul 6 03:40 unraid_video/ drwxrwxrwx 1 root root 4096 Jun 16 14:28 wd2t/ docker exec -it Tonido bash root@423b4d2af636:/# ls -al /unRAID/ total 8 drwxrwxrwx 6 root root 120 Jul 6 19:12 . drwxrwxrwx 1 root root 198 Jul 6 20:08 .. drwxrwxrwx 3 root root 4096 Dec 31 1969 BOOTUNRAID drwxrwxrwx 2 root root 0 Jul 6 03:40 unraid_unraid drwxrwxrwx 2 root root 0 Jul 6 03:40 unraid_video drwxrwxrwx 1 root root 4096 Jun 16 14:28 wd2t /usr/bin/docker run -d --name="Tonido" --net="bridge" -e TZ="America/Chicago" -p 10001:10001/tcp -v "/mnt/user/appdata/tonido":"/config":rw -v "/mnt/disks/":"/unRAID":ro captinsano/tonido

None of the docker veterans know how to do it? I experimented with symlinks and they are totally ignored by docker. I also tried to mount a directory with option bind (mount --bind /<array share>/<dir> /<cache disk/<dir>) and it seems to work i.e. such directory is visible for docker. One thing to note is that you probably should not bind two user shares like so: /mnt/user/download/jdownloader on /mnt/user/appdata/jdownloader/mydownloads type none (rw,bind) (Binding two user shares produces strange results) but direct reference to cache works /mnt/user/download/jdownloader on /mnt/cache/appdata/jdownloader/mydownloads type none (rw,bind) command: mount -o bind /mnt/user/download/jdownloader/ /mnt/cache/appdata/jdownloader/mydownloads/ So far the above seems to work (also mover behaves correctly as /mnt/user/download utilizes cache) Any comments?

I keep all my dockers /config on appdata cache-only share and it works fine in most cases. As cache disk capacity is rather limited is it safe to symlink some of the subdirectories of /config to array share for storing larger amounts of data? Example: JDownloader docker uses /config/Downloads for storing downloaded content. In some cases downloads can take lots of cache space. I'd like to symlink /config/Downloads (in fact /mnt/user/appdata/jdownloader/Downloads/) so it points to /mnt/user/<array sharename>/<dir>. Effectively JD config files would stay on cache, but downloaded content would be sent straight to disk array. Is it safe/effective way to handle the problem?

I second RobJ, some kind of two-stage approach would be beneficial. Probably 1st threshold to try to spin down and 2nd to unconditionally shut down.

It works now: I deleted appdata, recreated docker, mountpoints in Unassigned Devices and restarted unraid. After that it started working. I can either pass the actual mount and specify subdir in Ubooquity or pass mount+subdir. It works either way. At least that makes sense, I don't know what was wrong in the beginning. It got me thinking passing mountpoints is not allowed. I'm glad it is. Thanks Squid!

So you passthrough the "bare share", without directories following it /mnt/disks/server_b_EBooks/ vs /mnt/disks/unraid_share1/subdir1/subdir2 maybe that's the clou. Will test later....thanks for encouragement

I'm trying to pass a "mountpoint" /mnt/disks/192.168.0.150_unraid/eBooks/eBooks/ to the docker below as /mnt /usr/bin/docker run -d --name="Ubooquity" --net="bridge" -e TZ="America/Chicago" -p 8085:8085/tcp -v "/mnt/cache/appdata/ubooqquity/":"/config":rw -v "/mnt/disks/192.168.0.150_unraid/eBooks/eBooks/":"/mnt":rw hurricane/ubooquity Well, docker doesn't see the contents of said /mnt/disks/192.168.0.150_unraid/eBooks/eBooks/ I confirmed that docker's /mnt is empty using also docker exec bash, . There is no problem if I substitute that mountpoint with a local path. Actually /mnt/disks/192.168.0.150_unraid/eBooks/eBooks/ comes from 'Unassigned Devices' plugin and is a valid mountpoint for another unraid. I vaguely remember that using "remote directories" was a problem with docker long time ago. Is it still an issue, any way to circumvent this? Thanks.

Latest update (2015.06.15a) broke my export of USB drive to NFS. It is independent if I check or not 'run script in background'. And manual unmount followed by mount properly exports NFS i.e. script is executed. Tue Jun 16 0:10:34 CDT 2015: Drive found with the following attributes: serial='WD_Elements_10B8_575841314131353253354850-0:0', serial_short='575841314131353253354850', device='/dev/sdk1', part='1', label='Elements', fstype='ntfs', target='', size='2000364240896', used='0', avail='2000364240896', mountpoint='/mnt/disks/wd2t', owner='user', automount='1', shared='1', command='/boot/config/plugins/unassigned.devices/wd2t.sh', command_bg='false' Tue Jun 16 0:10:34 CDT 2015: Adding disk ... Tue Jun 16 0:10:34 CDT 2015: Mounting drive with command: mount -t ntfs -o auto,async,nodev,nosuid,umask=000 '/dev/sdk1' '/mnt/disks/wd2t' Tue Jun 16 0:10:43 CDT 2015: Successfully mounted '/dev/sdk1' on '/mnt/disks/wd2t' Tue Jun 16 0:10:43 CDT 2015: Defining share 'wd2t' on file '/etc/samba/unassigned-shares/wd2t.conf' . Tue Jun 16 0:10:43 CDT 2015: Adding share wd2t to /boot/config/smb-extra.conf Tue Jun 16 0:10:43 CDT 2015: Reloading Samba configuration. Tue Jun 16 0:10:43 CDT 2015: Directory '/mnt/disks/wd2t' shared successfully. Tue Jun 16 0:10:43 CDT 2015: Running command '/boot/config/plugins/unassigned.devices/wd2t.sh' with action 'ADD'. Tue Jun 16 0:10:43 CDT 2015: Disk Elements shared successfully. cat /tmp/WD_Elements_10B8_575841314131353253354850-0\:0.log exporting *:/mnt/disks/wd2t In case you can't figure it out can I revert to the previous code?

Just for reference, that seems to be a good case: Fri Jun 12 12:49:35 CDT 2015: Drive found with the following attributes: serial='WD_Elements_10B8_575841314131353253354850-0:0', serial_short='575841314131353253354850', device='/dev/sdg1', part='1', label='Elements', fstype='ntfs', target='', size='2000364240896', used='0', avail='2000364240896', mountpoint='/mnt/disks/wd2t', owner='user', automount='1', shared='1' Fri Jun 12 12:49:35 CDT 2015: Adding disk ... Fri Jun 12 12:49:35 CDT 2015: Mounting drive with command: mount -t ntfs -o auto,async,nodev,nosuid,umask=000 '/dev/sdg1' '/mnt/disks/wd2t' Fri Jun 12 12:49:44 CDT 2015: Successfully mounted '/dev/sdg1' on '/mnt/disks/wd2t' Fri Jun 12 12:49:44 CDT 2015: Defining share 'wd2t' on file '/etc/samba/unassigned-shares/wd2t.conf' . Fri Jun 12 12:49:44 CDT 2015: Adding share wd2t to /boot/config/smb-extra.conf Fri Jun 12 12:49:44 CDT 2015: Reloading Samba configuration. Fri Jun 12 12:49:44 CDT 2015: Directory '/mnt/disks/wd2t' shared successfully. Fri Jun 12 12:49:44 CDT 2015: Running command '/boot/config/plugins/unassigned.devices/wd2t.sh' with action 'ADD'. Fri Jun 12 12:49:44 CDT 2015: Disk Elements shared successfully

The plugin has been working very well, I managed to preclear more than 10 disks with it - no problems so far. Only cosmetic issue is that upon completion it displays "Preclear Successful^n Total TIme xxx^n ...etc". Basically there is an extra "^n" after each string. This has been happening on recent Chrome versions each time. Thanks for the plugin!

Happy to report that with latest update (.11a) your plugin executes my script to export a share under NFS. Thanks.

I have a sata drive formatted xfs that mounts and unmounts and initiates the script properly. I have a USB formatted ntfs that is mounting, but the script is not initiated. So it appears that something is not working in RC5. I'm on rc3 and it seems my script is not executed upon boot. I have to manually unmount and mount it again to get my script executed (script is related to exporting USB drive under NFS). I described my problem couple of posts back, but it seems your problem may be similar.

I added something like that (quick and dirty) as a script. It works when USB disk is mounted/unmounted manually from GUI. But it doesn't work when unRAID is restarted. It is not executed looking at /etc/exports contents. case $ACTION in 'ADD' ) // do your stuff here echo '"/mnt/disks/wd2t" -async,no_subtree_check,fsid=150 *(rw,insecure,anongid=100,anonuid=99,all_squash)' >>/etc/exports exportfs -rav ;; 'REMOVE' ) sed '/wd2t/d' /etc/exports > /tmp/exp cp -f /tmp/exp /etc/exports exportfs -rav ;; esac

Excellent plugin! Could you please also enable NFS sharing or provide script to mount/umount the device?