Niccarter Posted September 12, 2014 Share Posted September 12, 2014 This is really bad! I've upgraded from beta7 to beta9 and all my containers have gone. The documentation is all over the place - you have to go back to beta8 to change the docker configuration only to find the template directory has been moved back to where it was in beta7. The screenshots in threads to not match what you see in in beta9 I have started Docker and I understand that I need to redownload the docker files, but I can neither access my original templates nor access the new ones. Please fix this urgently. Regards Nc See screenshot.. Quote Link to comment
Trunkz Posted September 12, 2014 Share Posted September 12, 2014 I've just upgraded to beta9, and my two cache devices are showing as unformatted. This has happened twice now; the last time going from beta7 to 8. Quote Link to comment
PeterB Posted September 12, 2014 Share Posted September 12, 2014 I have started Docker and I understand that I need to redownload the docker files, but I can neither access my original templates nor access the new ones. I'm in a similar situation, but only started running docker apps in beta8, where all was great. Now, in beta9, although one docker app is able to run (deluge), my other docker container (LogitechMediaServer) has vanished and there are no containers listed/available to download. HELP!!!! Quote Link to comment
NAS Posted September 12, 2014 Share Posted September 12, 2014 In beta 8 the my* templates were in /var/lib/docker/unraid-templates i have to say it is not immediately clear how to and where to place everything as there are various conflicting posts Quote Link to comment
jonp Posted September 12, 2014 Share Posted September 12, 2014 All, we will work today to straighten out any confusion. Please bear in mind the #1 most important thing to us was to first get out this patch for REISERFS. Quote Link to comment
Squid Posted September 12, 2014 Share Posted September 12, 2014 All, we will work today to straighten out any confusion. Please bear in mind the #1 most important thing to us was to first get out this patch for REISERFS. Completely understandable. Going from 8 to 9 worked perfectly once I followed this: Also, maybe its just me, but I also have lost the ability to add docker containers via templates. The drop down menu just states "Select a Template". If I revert back to beta 8, all the templates are there, but they disappeared from my beta9 But, all of the containers which I was running before are still running. Sorry for that. We changed the user templates back to the flash drive: BETA 8 BETA 9 /var/lib/docker/unraid-templates /boot/config/plugins/dockerMan/templates-user While LT don't create their own containers, you can use several of community ones using the templates below. [glow=red,2,300]UPDATE:[/glow] To help the transition, I've created a template repository here: https://github.com/gfjardim/docker-containers/tree/templates After update the plugin, a new field will appear on the Docker extension page. Add this address there and click Save. Quote Link to comment
bkastner Posted September 12, 2014 Share Posted September 12, 2014 Maybe its just a stupid question, but I see that when the array is stopped that I now have the option to change the number of slots... Why would I ever want to do that? That feature was added in 6.0 beta 7. Use it to create a BTRFS cache pool with multiple drives for fault tolerance. Older versions only allowed a single cache drive, but now you can have multiple drives in a cache pool. Also, if you have a Pro key, but only have a few drives it means when you stop the array you can select it to only show 5-6 slots (or your drive number +1) instead of all 23. Just makes the GUI cleaner. Quote Link to comment
SSD Posted September 12, 2014 Share Posted September 12, 2014 Upgrading from 6b7 to 6b9 with Docker on non-cache SSD My array was already stopped, although powered up. Your starting place may be different. You can do this in a very similar way by loading the flash into your workstation. Before upgrading (still booted into 6b7 with array stopped), I deleted the dockerman.plg (actually I just renamed it to ".old") in the \\tower\flash\config\plugins directory. I then made a new directory \\tower\flash\config\plugins\dockerMan\templates-user I then copied the templates from \\tower\flash\config\plugins\Docker to \\tower\flash\config\plugins\dockerMan\templates-user (I didn't delete the old ones yet). I then copied the beta9 files over (bzroot, bzimage, xen, readme.txt) to the root of the flash (//tower/flash) as per Tom's instructions. I am one that has an SSD for dockers and not using the loopback file system (although I may switch over as I'm wasting a lot of SSD space!). I removed the mount related commands from my go file (see the "#" added to front of the two lines to comment them out). #mkdir -p /mnt/btrfs #mount /dev/disk/by-id/ata-Samsung_SSD_840_EVO_250GB_S1DBNSCF31xxxxx-part1 /mnt/btrfs I then rebooted (interesting but the Web GUI was unresponsive and could reboot that way, so I went to the unmenu directory and ran the power down script (99-unmenu_user_script_powerdown) which shut it down. I then powered up. I wonder if removing the dockerman plugin confused the GUI.) unRAID 6b9 came up without any problem. Started array (mine defaults to not start). All smooth. I went to the Docker page and added a reference to my SSD in the docker image field (/dev/disk/by-id/ata-Samsung_SSD_840_EVO_250GB_S1DBNSCF3xxxxx-part1). unRAID seemed happy with it and mounted it properly. Under the mount I saw this status: Label: none uuid: f79f2dd9-6838-4362-a915-36093edxxxxx Total devices 1 FS bytes used 2.87GiB devid 1 size 232.88GiB used 6.02GiB path /dev/sdg1 Btrfs v3.16 (Actually I didn't grab the exact content at this time. What you see above is what the field says after completing the remaining steps below. I think that because my Docker containers were re-downloaded, I am probably using 2x as much space as I should be.) Going to the docker page, I found that my containers were not showing (I expected them to be there). I was able to click the "+" icon and see that the "my-" templates were there thanks to my copying before rebooting). I added the standard templates directory (https://github.com/gfjardim/docker-containers/tree/templates) to the "Templates repositories:" field and clicked "Save". Now clicking the + allowed me to see both my templates and the standard templates. From this screen I selected my first Docker template. From its configuration screen, I clicked "Add" at the bottom right, and saw the processing dialog. It said that it could not find the image file and began to download it. In the past, despite the dialog hiding part of the Docker screen, I was able to see the status of the download on the screen beneath. Not so this time. The dialog said to wait and I was patient and waited for it to complete. Took a minute or two and it said it was completed successfully. The Docker had stated automatically. I did the same for my other Docker. There were no other changes I needed to make. Both my dockers started and are working fine with no configuration changes. (Note that my appdata folders are not on my ssd - they are on my cache disk). If yours are on you SSD, you'll have some changes to make since the mount point to the SSD will have changed to "/var/lib/docker". I did some hunting around looking at newly added files on my RFS disks. Opening text files, looking a images, etc. I saw no signs of corruption after about 10 minutes of random checking. I do have pre-beta7 md5 files that I plan to use to verify my data, and will make a followup post with my findings in a day or two. Quote Link to comment
SSD Posted September 12, 2014 Share Posted September 12, 2014 Upgrading from 6b7 to 6b9 with Docker on non-cache SSD ... Label: none uuid: f79f2dd9-6838-4362-a915-36093edxxxxx Total devices 1 FS bytes used 2.87GiB devid 1 size 232.88GiB used 6.02GiB path /dev/sdg1 Btrfs v3.16 (Actually I didn't grab the exact content at this time. What you see above is what the field says after completing the remaining steps below. I think that because my Docker containers were re-downloaded, I am probably using 2x as much space as I should be.) ... Going to the /var/lib/docker mount point, I see the following in the root .... total 28 drwxr-xr-x 1 root root 192 Sep 12 06:10 ./ drwxr-xr-x 12 root root 240 Sep 11 22:22 ../ drwx------ 1 root root 20 Sep 12 06:03 btrfs/ drwx------ 1 root root 256 Sep 12 06:10 containers/ drwxrwxrwx 1 root root 148 Aug 16 10:51 docker/ drwx------ 1 root root 12 Sep 12 05:59 execdriver/ drwx------ 1 root root 7176 Sep 12 06:10 graph/ drwx------ 1 root root 32 Sep 12 05:59 init/ -rw-r--r-- 1 root root 5120 Sep 12 06:10 linkgraph.db -rw------- 1 root root 208 Sep 12 06:10 repositories-btrfs drwx------ 1 root root 0 Sep 12 05:59 tmp/ drwxrwxrwx 1 root root 0 Sep 12 05:59 unraid-templates/ drwx------ 1 root root 0 Sep 12 05:59 volumes/ Tom - is it safe to delete and remove the "docker" folder? If so, can I just run the command "rm -r /var/lib/docker/docker"? Quote Link to comment
TSM Posted September 12, 2014 Share Posted September 12, 2014 I've just upgraded to beta9, and my two cache devices are showing as unformatted. This has happened twice now; the last time going from beta7 to 8. This worries me. I was about to upgrade from Beta7 to Beta9. I have btrfs formatted mirrored ssd cache drives. Because of the current problems, I haven't let mover run, and I've got about 60gigs of files on my cache. Will the files and the formatting of the cache still be in place after upgrading? Quote Link to comment
jjslegacy Posted September 12, 2014 Share Posted September 12, 2014 No issues here going from Beta8 to Beta9. Copied my docker templates and added the template repository. BTRFS cache drive came back fine. Doing some testing now but so far so good. Hopefully this is behind us. Great job guys. Quote Link to comment
dtempleton Posted September 12, 2014 Share Posted September 12, 2014 my* templates: root@Tower:~# mkdir /boot/config/plugins/dockerMan/templates-user/ root@Tower:~# cp -a /var/lib/docker/unraid-templates/* /boot/config/plugins/dockerMan/templates-user/ root@Tower:~# v /boot/config/plugins/dockerMan/templates-user/ total 40 drwxrwxrwx 2 root root 4096 Sep 12 10:38 ./ drwxrwxrwx 3 root root 4096 Sep 12 10:38 ../ -rwxrwxrwx 1 root root 868 Sep 11 12:41 my-CouchPotato.xml* -rwxrwxrwx 1 root root 639 Sep 6 12:15 my-Dropbox.xml* -rwxrwxrwx 1 root root 657 Sep 6 20:57 my-HTPC-Manager.xml* -rwxrwxrwx 1 root root 631 Sep 6 15:36 my-MariaDB.xml* -rwxrwxrwx 1 root root 772 Sep 3 21:42 my-SABnzbd.xml* -rwxrwxrwx 1 root root 1035 Sep 3 23:51 my-SickBeard.xml* -rwxrwxrwx 1 root root 821 Sep 3 23:21 my-XBMCServer.xml* -rwxrwxrwx 1 root root 897 Sep 6 16:39 my-ownCloud.xml* Seems straightforward. More worried about the unformatting cache drive. Quote Link to comment
TSM Posted September 12, 2014 Share Posted September 12, 2014 I've just upgraded to beta9, and my two cache devices are showing as unformatted. This has happened twice now; the last time going from beta7 to 8. This worries me. I was about to upgrade from Beta7 to Beta9. I have btrfs formatted mirrored ssd cache drives. Because of the current problems, I haven't let mover run, and I've got about 60gigs of files on my cache. Will the files and the formatting of the cache still be in place after upgrading? Said screw it. Copied the contents of my cache drive to my google drive. Got a free 100gigs when I bought my chromebook 8 months ago. Upgrade from Beta7 to Beta9 seems to have gone smooth. Cache drives look normal. Ran mover manually seems fine. Then reset it back to nightly schedule. Your mileage may vary, f**k if I know if it will work as well for you. Quote Link to comment
limetech Posted September 12, 2014 Author Share Posted September 12, 2014 @LT What is the safest/best format to transfer to? BTRFS or XFS? or maybe wait for EXT4? I know that are discussion on several places about this. But I need your opinion on this. BR Peter I mentioned somewhere else that we are using xfs now for all AVS-10/4 server builds. Here's an article on the subject from Redhat: http://rhelblog.redhat.com/2014/01/07/best-file-system/ Another: http://www.eweek.com/enterprise-apps/red-hat-releases-rhel-7-with-new-filesystem-docker-support.html Quote Link to comment
jonp Posted September 12, 2014 Share Posted September 12, 2014 Upgrading from 6b7 to 6b9 with Docker on non-cache SSD ... Label: none uuid: f79f2dd9-6838-4362-a915-36093edxxxxx Total devices 1 FS bytes used 2.87GiB devid 1 size 232.88GiB used 6.02GiB path /dev/sdg1 Btrfs v3.16 (Actually I didn't grab the exact content at this time. What you see above is what the field says after completing the remaining steps below. I think that because my Docker containers were re-downloaded, I am probably using 2x as much space as I should be.) ... Going to the /var/lib/docker mount point, I see the following in the root .... total 28 drwxr-xr-x 1 root root 192 Sep 12 06:10 ./ drwxr-xr-x 12 root root 240 Sep 11 22:22 ../ drwx------ 1 root root 20 Sep 12 06:03 btrfs/ drwx------ 1 root root 256 Sep 12 06:10 containers/ drwxrwxrwx 1 root root 148 Aug 16 10:51 docker/ drwx------ 1 root root 12 Sep 12 05:59 execdriver/ drwx------ 1 root root 7176 Sep 12 06:10 graph/ drwx------ 1 root root 32 Sep 12 05:59 init/ -rw-r--r-- 1 root root 5120 Sep 12 06:10 linkgraph.db -rw------- 1 root root 208 Sep 12 06:10 repositories-btrfs drwx------ 1 root root 0 Sep 12 05:59 tmp/ drwxrwxrwx 1 root root 0 Sep 12 05:59 unraid-templates/ drwx------ 1 root root 0 Sep 12 05:59 volumes/ Tom - is it safe to delete and remove the "docker" folder? If so, can I just run the command "rm -r /var/lib/docker/docker"? Yes, you can remove the docker folder. Not sure why you have a docker sub-folder off there though. Here's my directory tree from my /var/lib/docker: root@newtower:/var/lib/docker# v total 20 -rw-rw-rw- 1 root root 81 Aug 31 01:53 autostart drwx------ 1 root root 20 Aug 30 08:27 btrfs/ drwxrwxrwx 1 root root 0 Sep 4 23:44 community-templates/ drwx------ 1 root root 1664 Sep 8 23:00 containers/ drwx------ 1 root root 12 Aug 30 08:18 execdriver/ drwx------ 1 root root 34440 Sep 7 01:04 graph/ drwxrwxrwx 1 root root 20 Sep 4 23:44 images/ drwx------ 1 root root 32 Aug 30 08:18 init/ -rw-r--r-- 1 root root 7168 Sep 8 23:00 linkgraph.db -rw------- 1 root root 1315 Sep 8 23:00 repositories-btrfs drwxrwxrwx 1 root root 650 Sep 1 13:09 templates/ drwx------ 1 root root 0 Aug 30 08:18 tmp/ drwxrwxrwx 1 root root 12 Sep 7 12:13 unraid/ -rw-rw-rw- 1 root root 51 Sep 4 08:43 unraid-autostart drwxrwxrwx 1 root root 52 Sep 7 01:01 unraid-templates/ drwx------ 1 root root 6 Aug 30 08:54 vfs/ drwx------ 1 root root 392 Aug 30 08:54 volumes/ Here's what I would do though just to make sure you don't have any btrfs subvolumes accidentally stored under the extra docker folder. 1) From command line, navigate to /var/lib/docker/docker/btrfs/subvolumes 2) Print out a listing of the directory (use "ls" or "v") 3a) If there are no sub-directories present, skip to step 4. 3b) If there ARE sub-directories present, use the following command to remove them (while navigated under /var/lib/docker/docker/btrfs/subvolumes): btrfs subvolume delete * 4) Navigate back out to /var/lib/docker and issue a "rm -r docker" command. Quote Link to comment
dlandon Posted September 12, 2014 Share Posted September 12, 2014 I'm not sure where to report this potential bug in dockerMan. I create my dockers just fine, but they cannot be edited. It comes up with a blank page when I try to edit. The my-* templates have been moved to the flash, but I think the edit is still pointing to the old location. I can see the templates on the flash at /boot/config/plugins/dockerMan/templates-user. I can delete the container and then re-create it from the my-* template just fine. Quote Link to comment
BRiT Posted September 12, 2014 Share Posted September 12, 2014 I have experienced the same trouble. The edit page shows with no values at all. To make modifications one has to drop down to manually editing the template file or deleting the docker and recreating from the same template with your changes. Quote Link to comment
ogi Posted September 12, 2014 Share Posted September 12, 2014 stuff... ... stuff... Yes, you can remove the docker folder. Not sure why you have a docker sub-folder off there though. Here's my directory tree from my /var/lib/docker: root@newtower:/var/lib/docker# v total 20 -rw-rw-rw- 1 root root 81 Aug 31 01:53 autostart drwx------ 1 root root 20 Aug 30 08:27 btrfs/ drwxrwxrwx 1 root root 0 Sep 4 23:44 community-templates/ drwx------ 1 root root 1664 Sep 8 23:00 containers/ drwx------ 1 root root 12 Aug 30 08:18 execdriver/ drwx------ 1 root root 34440 Sep 7 01:04 graph/ drwxrwxrwx 1 root root 20 Sep 4 23:44 images/ drwx------ 1 root root 32 Aug 30 08:18 init/ -rw-r--r-- 1 root root 7168 Sep 8 23:00 linkgraph.db -rw------- 1 root root 1315 Sep 8 23:00 repositories-btrfs drwxrwxrwx 1 root root 650 Sep 1 13:09 templates/ drwx------ 1 root root 0 Aug 30 08:18 tmp/ drwxrwxrwx 1 root root 12 Sep 7 12:13 unraid/ -rw-rw-rw- 1 root root 51 Sep 4 08:43 unraid-autostart drwxrwxrwx 1 root root 52 Sep 7 01:01 unraid-templates/ drwx------ 1 root root 6 Aug 30 08:54 vfs/ drwx------ 1 root root 392 Aug 30 08:54 volumes/ Here's what I would do though just to make sure you don't have any btrfs subvolumes accidentally stored under the extra docker folder. 1) From command line, navigate to /var/lib/docker/docker/btrfs/subvolumes 2) Print out a listing of the directory (use "ls" or "v") 3a) If there are no sub-directories present, skip to step 4. 3b) If there ARE sub-directories present, use the following command to remove them (while navigated under /var/lib/docker/docker/btrfs/subvolumes): btrfs subvolume delete * 4) Navigate back out to /var/lib/docker and issue a "rm -r docker" command. Jonp, I can't speak for all new people here, but personally, I love how you are sharing how your personal setup is. In another thread you posted your array drives are ReiserFS (when asked about your own setup). As a linux noob, seeing all the possible options can be overwhelming, and often I don't feel like I have the vocabulary to fully comprehend the effects of all these options, but you describing your own setup gives me a fantastic point to start from. I appreciate that, and I'm sure most linux noobs do as well. Ogi Quote Link to comment
jonp Posted September 12, 2014 Share Posted September 12, 2014 Seems there are a number of folks having struggles with Docker and beta9. Would it be helpful if we created an IRC channel to temporarily help folks who need it? Is there another medium that folks would prefer to use? I want to make sue everyone gets the help they need to get to beta 9 and back up and running with all their containers and templates intact. Quote Link to comment
dlandon Posted September 12, 2014 Share Posted September 12, 2014 Coming from beta8, I haven't had any problems except the template editing bug I reported earlier. I think the people that are having the biggest problems are those who set up the docker.img on a btrfs disk and now are trying to convert to the loop file for docker. Quote Link to comment
fiscalcon Posted September 12, 2014 Share Posted September 12, 2014 I'm not sure where to report this potential bug in dockerMan. I create my dockers just fine, but they cannot be edited. It comes up with a blank page when I try to edit. The my-* templates have been moved to the flash, but I think the edit is still pointing to the old location. I can see the templates on the flash at /boot/config/plugins/dockerMan/templates-user. I can delete the container and then re-create it from the my-* template just fine. Experiencing the same thing. Thought it was causing my NZBDrone docker to not work correctly (couldn't manually search for episodes anymore), but that ended up being something in my /appdata folder. Never found out what, just wiped and started over and the issue went away. Quote Link to comment
NAS Posted September 12, 2014 Share Posted September 12, 2014 Seems there are a number of folks having struggles with Docker and beta9. Would it be helpful if we created an IRC channel to temporarily help folks who need it? Is there another medium that folks would prefer to use? I want to make sue everyone gets the help they need to get to beta 9 and back up and running with all their containers and templates intact. Or join the existing one? Quote Link to comment
limetech Posted September 12, 2014 Author Share Posted September 12, 2014 I'm not sure where to report this potential bug in dockerMan. I create my dockers just fine, but they cannot be edited. It comes up with a blank page when I try to edit. The my-* templates have been moved to the flash, but I think the edit is still pointing to the old location. I can see the templates on the flash at /boot/config/plugins/dockerMan/templates-user. I can delete the container and then re-create it from the my-* template just fine. Yes I see the problem & will post a fix. Quote Link to comment
clowrym Posted September 12, 2014 Share Posted September 12, 2014 I had no issues with upgrading / Docker, all is working great w/ the instructions here!! Quote Link to comment
limetech Posted September 12, 2014 Author Share Posted September 12, 2014 Ok, let's give this a whirl: Paste this URL into the "Install Extension" box on the Extensions page and then click Install. This will download dockerMan with the my-* edit bug fixed. https://raw.githubusercontent.com/limetech/dockerMan/master/dockerMan.plg Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.