aptalca

To be honest, I never tested usb devices with any dockers. This container is running a minimal ubuntu (phusion) and usb support should be in there, but I can't be certain.

See the two answers on the following page: http://stackoverflow.com/questions/24225647/docker-any-way-to-give-access-to-host-usb-or-serial-device You can add the parameters ("--device=/dev/blah" or "-v /dev/bus/usb:/dev/bus/usb" in "Extra Parameters" in "Advanced View" while editing the container in unraid gui

I'll look into the username and password options. By the way, the rdp-calibre docker (a separate one) has the url-prefix option and includes the calibre server

1. Use a docker for dynamic dns. There are several different ones and they ask work great 2. Routers usually have limited processing power and ram. Openvpn running on unraid is much much faster. Openvpn running on a router, i was getting 400kbps data transfer speed. With openvpn in docker, I'm getting more than 5000kbps 3. You always have forward a port for vpn (unless your gateway is running it) and that's considered safe as long as that is the only port open

I'm assuming this is plexrequests? What is the path you used for the config folder? I've been helping jrdnlc with this issue. (teamviewer) It's weird, and I'm at a loss to figure it out now. Steps taken. Old container & appdata folder deleted. New template, branch=dev. Pulls from docker hub fine, then starts as the logs shows (incidentally I think jrdnlc has posted a log from his more recent attempts at installing branch=master) There doesn't appear to be the pull from github like I see when I install on my machine. I've deleted the container & image multiple times as well as deleted the appdata folder using midnight commander. Restarted the machine and restarted dockerman. We must be missing something really simple but I can't for the life of me see what. I've even uploaded my .xml file to my github repo and tried using that even though the information contained within is no different. EDIT: Also unable to reinstall master branch, and I have reinstalled mine several times without issue throughout all this. Also, you're right, I didn't read the change logs, my bad.... Make sure that the app folder path is not /mnt/user/blah but instead /mnt/cache/blah The only other thing I can think of is issues with updating/installing meteor because that is the next step in that install process

I'm assuming this is plexrequests? What is the path you used for the config folder?

Try running the dev branch, a fix for ssl was merged in May that hasn't made it to the Master yet. https://github.com/lokenx/plexrequests-meteor/pull/82 That worked, although for anyone else reading this, the instructions to put EDGE=1 in the Environmental Variables don't work. Instead add an Environmental Variable BRANCH = dev EDIT: My docker template must be very old.... Sorry aptalca... Looks like someone didn't read the change log [emoji6] I hope the issue of template xmls not being updated gets addressed in dockerman

What do the logs say?

Its a bug within dockerMan, not CA. You're going to have to basically delete all references to the template file within /boot/config/plugins/dockerMan. See here http://lime-technology.com/forum/index.php?topic=40262.msg380740#msg380740 i have deleted the files that is related to cadvisor: /boot/config/plugins/dockerMan/templates-user/my-cadvisor.xml /boot/config/plugins/dockerMan/images/ 2 image files related to cadvisor delete container and the image reinstall cadvisor but still no webui link do i miss smtg? Reboot

Click on the advanced view (top right). There are more fields to fill in there Thank you, I don't know why I thought the advanced settings were entered after the docker was created. I feel extra dumb now. [emoji6] I'm glad you got it working

Hmm, that might be a question best suited for the plexrequests thread on plex. Sorry I can't help you there

Click on the advanced view (top right). There are more fields to fill in there

Can you elaborate on that? A few of my dockers have the same problem and I'd love to know how to fix it. Thanks I´m not 100% sure but it´s only on unRAID this issue exist tried on ubuntu, debian(omv) and centos with a volume on an nfs share. The issue was for me that i had a directory "/config/etc/socks" (that is symlinked inside the docker) and when openvpn starts it creates some files in the folder and when ever I used /mnt/user/ it couldent write to it (I could however create files manually on host and inside the docker). If I used /mnt/disk/ it would work, so I then noticed that while doing this if I exexuted a simple "ls -la" on the share /mnt/user/appdata/openvpn/socks and it would just output that the files dident exists and question marks on the user permissions. So i did the simple thing since the files in question here dident need permanent storage and I moved them to a /tmp folder inside the container. But the only thing I can think of that is causing this is that the /mnt/user share runs through samba? and has an issue with traversing symlinks. Hmm that makes sense. Quite a few of my dockers use symlinks inside the config folder. I'll test that with my dockers too. Thanks so much for the report

Can you elaborate on that? A few of my dockers have the same problem and I'd love to know how to fix it. Thanks

It's the server version. I left the Container Volume as /config and my Host Path is /mnt/user/Programs/unRAID/appdata/Calibre/ I haven't changed libraryinternalpath If the host path is a cache only share, can you try changing that to /mnt/cache/Programs/unRAID/appdata/Calibre/ ? Also, since it's the server version, did you have an existing calibre library at that location or did you add books through command line? If there was an existing library, do the calibre versions of the library creator and this new server match?

I just did a fresh install with the dev branch and within a couple of minutes everything was installed and it was fired up. Here's what I see in my log: First install detected, cloning repository Using the dev branch Branch dev set up to track remote branch dev from origin. npm-container: updating npm dependencies -- git-rev-sync... Sep 23 14:09:46 cc54b1595e56 syslog-ng[123]: syslog-ng starting up; version='3.5.3' [[[[[ /config/plexrequests-meteor ]]]]] => Started proxy. => Meteor 1.2.0.1: React and Angular support, ECMAScript 2015, new faster build system, and improvements to the Cordova integration for mobile apps. This release is being downloaded in the background. Update your app to Meteor 1.2.0.1 by running 'meteor update'. => Started MongoDB. => Started your app. => App running at: http://localhost:3000/ [34mI20150923-14:10:02.876(-4) (percolate_synced-cron.js:85) [39mSyncedCron: Scheduled "Update CP download status" next run @Wed Sep 23 2015 18:00:00 GMT-0400 (EDT) See in the log where it says "Started your app" under "Started MongoDB". Do you not get that message? Could you perhaps be trying to use a port that is already being used by something else? Yeah i dont seem to get that message. Or the 'app running at'. However i do get the cronjobs all firing off successfully. I've tried reinstalling to a different location - And that works fine, so it seems that somethngs gone wrong with the DB thats stopping it from launching. Any ideas on what i can do? PITA having to set it all up again so trying to avoid that if possible In that case, you can perhaps ask on their plex forums. I'm not really an expert on the inner workings of the app. I just know how to set it up and make it run in docker.

If you stop/restart it won't have to download the update again. However if you remove/reinstall the container, then it will have to download again. These containers really have two parts to them: 1) The OS+app portion: resides in the docker image file. The specific OS image is downloaded from the dockerhub and it is static (in the calibre's case the image downloaded contains the ubuntu/phusion base and an older but stable version of calibre already installed) Any changes made to this OS+app portion are lost if container is removed. When you reinstall, you go back to the static image. 2) Local data portion: resides in the folder of your choice. Any files that we want to be persistent, such as app data, config files, library, etc. we keep in this external partition. When you reinstall the container, as long as you select the same local data folder, your data will be persistent. In this case, when calibre updates itself, it is essentially updating the OS+app portion. If you remove or reinstall the container the OS+app portion is wiped clean and a fresh image is downloaded. But if you just stop and restart it uses the existing image file with all of your changes.

What does the log tell you? With edge set to 1, it checks for an update upon start and if there is a new update it will download and install it before guacamole is fired up. Unfortunately calibre's server can be super slow for downloading the package. Usually it takes a few minutes but one time it took me over half an hour to download it. The log should show the download progress and the percentage. I just restarted mine and let it update to the latest. Once complete it works fine. Ok, I assumed once the docker started that the app was updated/installed I'll start it with edge and give it a good 30 minutes just in case. (Their website is always slow for downloads). Without edge, it is already installed and ready to go, but with edge it checks for updates on start In the unraid webgui, under the docker tab, if you hit the log button on the right hand side, it will show you the download percentage in real time

I just did a fresh install with the dev branch and within a couple of minutes everything was installed and it was fired up. Here's what I see in my log: First install detected, cloning repository Using the dev branch Branch dev set up to track remote branch dev from origin. npm-container: updating npm dependencies -- git-rev-sync... Sep 23 14:09:46 cc54b1595e56 syslog-ng[123]: syslog-ng starting up; version='3.5.3' [[[[[ /config/plexrequests-meteor ]]]]] => Started proxy. => Meteor 1.2.0.1: React and Angular support, ECMAScript 2015, new faster build system, and improvements to the Cordova integration for mobile apps. This release is being downloaded in the background. Update your app to Meteor 1.2.0.1 by running 'meteor update'. => Started MongoDB. => Started your app. => App running at: http://localhost:3000/ [34mI20150923-14:10:02.876(-4) (percolate_synced-cron.js:85) [39mSyncedCron: Scheduled "Update CP download status" next run @Wed Sep 23 2015 18:00:00 GMT-0400 (EDT) See in the log where it says "Started your app" under "Started MongoDB". Do you not get that message? Could you perhaps be trying to use a port that is already being used by something else?

I've tried all the different branches incase that was causing the issue - Is it the fact meteor seems to have an update? Not sure what to do at this point. The log seems fine. After starting the mongodb, it will take another 10-20 second to start the app (may be longer but that's how long it takes for me). Did you wait long enough? Does it give an error later? Meteor update should not have an effect. I tested it again and it will download the update in the background and then it will keep saying the update is ready to be applied. You can leave it as is and it is harmless. Unfortunately, the update is not compatible with plexrequests at the moment so just leave it as is. In my case (I tried the master branch) it works just fine once the app loads. I also installed from scratch and again it worked fine. I'll try other branches later to see if they cause any issues. Hmm.. I have left it running for days, and it still doesn't work. Afterwards it just says the cron.hourly logs so it all seems to 'think' its running fine. Is there a log file somewhere i could check besides the one found in the docker page of unraid? Looking through the apps files i don't see one Where do you keep your configuration folder? Is it on a cache drive? If so, did you enter the location as /mnt/user/blah or /mnt/cache/blah ?

Is that the rdp version? What location do you have your library set to in calibre settings? If it's not /config, did you set the libraryinternalpath variable to the new location?

What does the log tell you? With edge set to 1, it checks for an update upon start and if there is a new update it will download and install it before guacamole is fired up. Unfortunately calibre's server can be super slow for downloading the package. Usually it takes a few minutes but one time it took me over half an hour to download it. The log should show the download progress and the percentage. I just restarted mine and let it update to the latest. Once complete it works fine.

I've tried all the different branches incase that was causing the issue - Is it the fact meteor seems to have an update? Not sure what to do at this point. The log seems fine. After starting the mongodb, it will take another 10-20 second to start the app (may be longer but that's how long it takes for me). Did you wait long enough? Does it give an error later? Meteor update should not have an effect. I tested it again and it will download the update in the background and then it will keep saying the update is ready to be applied. You can leave it as is and it is harmless. Unfortunately, the update is not compatible with plexrequests at the moment so just leave it as is. In my case (I tried the master branch) it works just fine once the app loads. I also installed from scratch and again it worked fine. I'll try other branches later to see if they cause any issues.

I've tried all the different branches incase that was causing the issue - Is it the fact meteor seems to have an update? Not sure what to do at this point. Could be. I'll look into it

I'm glad you figured out the new permissions, that was going to be my suggestion and thanks for sharing the go file command for others' reference. I tried to come up with ways to fix the docker container but couldn't think of a good method as it allows users to pick any location as the download location. Giving the user the ability to change the umask would complicate things and it wouldn't solve the issue of other smb users not being able to access. I think your solution is the best way to go about it until JDownloader2 adds the ability to set downloaded file permissions.