Nextcloud with Letsencrypt using OMV and docker-compose - Q&A

  • For those who would like to follw the roue of bind mounting the docker directory:


    1. Stop the docker daemon
    2. Chose a directory where you future docker root should be /path_to_my_new_docker_root
    3. Make sure /path_to_my_new_docker_root is not on a mergerfs
    4. for testing purposes: do a bind mount from /var/lob/docker to /path_to_my_new_docker_root (testing only)
    5. Start your docker deamon and check docker image ls -> no image should be found!
    6. Stop the docker daemon
    7. unmount /var/lib/docker again
    8. Check the contents of /path_to_my_new_docker_root -> the docker files should be there
    9. now remove all files in /path_to_my_new_docker_root
    10. Now you are prepared for the real thing
    11. Stop the docker daemon again
    12. Move the content of /var/lib/docker to /path_to_my_new_docker_root (or do a rsync -a ... to keep the old stuff)
    13. append a line to /etc/fstab
      /path_to_my_new_docker_root /var/lib/docker none bind 0 0
    14. Start the docker daemon and check docker image ls and docker volume ls and docker ps -a
      All your stuff should be there and the containers be running.
    15. Reboot and check again
    16. If everything works click the link in my signature :)

    If you got help in the forum and want to give something back to the project click here (omv) or here (scroll down) (plugins) and write up your solution for others.

  • But booting from the other disk is the only thing which makes sense when looking at the errors you posted.

    Files changing on disk after booting ....

    If you got help in the forum and want to give something back to the project click here (omv) or here (scroll down) (plugins) and write up your solution for others.

    • Offizieller Beitrag

    For those who would like to follw the roue of bind mounting the docker directory:


    1. Stop the docker daemon
    2. Chose a directory where you future docker root should be /path_to_my_new_docker_root
    3. Make sure /path_to_my_new_docker_root is not on a mergerfs
    4. for testing purposes: do a bind mount from /var/lob/docker to /path_to_my_new_docker_root (testing only)
    5. Start your docker deamon and check docker image ls -> no image should be found!
    6. Stop the docker daemon
    7. unmount /var/lib/docker again
    8. Check the contents of /path_to_my_new_docker_root -> the docker files should be there
    9. now remove all files in /path_to_my_new_docker_root
    10. Now you are prepared for the real thing
    11. Stop the docker daemon again
    12. Move the content of /var/lib/docker to /path_to_my_new_docker_root (or do a rsync -a ... to keep the old stuff)
    13. append a line to /etc/fstab
      /path_to_my_new_docker_root /var/lib/docker none bind 0 0
    14. Start the docker daemon and check docker image ls and docker volume ls and docker ps -a
      All your stuff should be there and the containers be running.
    15. Reboot and check again
    16. If everything works click the link in my signature :)

    Basically the same as I suggested.. which worked until the second OS drive came into play.

  • Thank you for the detailed info not sure what you are trying to get but this paypal link is not even working:P:):)

    https://www.paypal.com/donate?…Q0aIMjVZDxJaD8A2HZ7OdwrT-

    Take this one instead: https://www.openmediavault.org/?page_id=1149

    If you got help in the forum and want to give something back to the project click here (omv) or here (scroll down) (plugins) and write up your solution for others.

  • It was not a so called issue except it was impossible to have doker on a data drive and nvidia working since if you were moving the docker thru the usual way you had to changed the docker etc/docker /daemon.json to

    {

    "runtimes": {

    "nvidia": {

    "path": "/usr/bin/nvidia-container-runtime",


    "runtimeArgs": []

    }

    },

    "default-runtime": "nvidia"

    }


    And in this case doker went back to var lib

    Although right now the nividia runtime is not working with OMV 6 except there is a way to make it work Nvidia working with OMV 6

    Yes it is resolved

    :)

  • LOL.. You guys are gonna love Starbucks problem... we are finishing it up now... I don't think we'd have ever figured htis out (or I wouldn't have.. he wasn't getting a key)

    I am documenting everything, screenshots and so on. How can I share my guide? Directly here?

    One final thing for you guys running duckdns. IF YOU DO NOT HAVE A STATIC PUBLIC IP ADDRESS... (or even if you do, this won't hurt)... use this compose file in a stack and adjust as needed. This will basically allow your duckdns profile to be automatically updated if your public IP ever changes... thus keeping your services online. Just adjust the lines with the # and then remove that #


    following your guide, subdomains=wildcard, right?

  • I am documenting everything, screenshots and so on. How can I share my guide? Directly here?

    following your guide, subdomains=wildcard, right?

    If you mean on the duckdns YML quoted, NO.


    It is the subdomain you registered on duckdns:


    Code
    SUBDOMAINS=mysite.duckdns.org
    • Offizieller Beitrag

    I am documenting everything, screenshots and so on. How can I share my guide? Directly here?

    following your guide, subdomains=wildcard, right?

    No, that is only if you're using duckdns.... (note that is a duckdns contaiiner).. so that needs to be your subdomain. That container is so your IP will be updated if it changes.

    In swag.. the "wildcard" if you use duckdns.. isn't your duckdns subdomain (that is defined in the URL= of swag).. it's technically whatever service you are using (nextcloud.duckdns... airsonic.duckdns... etc.)

    • Offizieller Beitrag

    Starbuck1973 I'll try to figure out a way to make that more clear tomorrow.. in the duckdns container, it is JUST your subdomain you registered w/ duckdns.. not the full URL.


    swag and duckdns are completely different containers in how they deal w/ subdomains. The duckdns container isn't even 100% required.. it's just if your IP changes.. and it isn't updated in the IP on your duckdns management page.. all your services will go down until you manually update it. This container, just manually makes sure that the IP of your server and the IP address pointing at your duckdns domain.. stay the same.

    • Offizieller Beitrag

    I am documenting everything, screenshots and so on. How can I share my guide? Directly here?

    following your guide, subdomains=wildcard, right?

    By the way "your guide" is exactly what I put there in that thread.. your biggest problem.. was your github login to duckdns was not allowing swag to fetch a cert. As I think we finally figured out.. the swag container will only grant a cert if you're using a gmail login to duckdns... as once you logged in to duckdns w/ your gmail account and set up a subdomain, you got a cert no problem and the rest of that guide was fine.

  • Hello macom , did you can write you last edit in your first post?



    I'm only find "Aug 10th 2019".


    Thank you, for your nice work.



    Best regards

    openmedianer

    • Offizieller Beitrag

    What are you referring to? To my knowledge he's the only one who's ever edited that post


  • Last edited (Dec 30th 2021) so it still is recently edited.

    Code
    Edited 21 times, last by macom: corrected "depends on" (mariadb); put the correct names of the fields on the Welcome page of Nextcloud to specify the MySQL/MariaDB added Q&A link removed exposed ports of nextcloud as not needed added SWAG replaced letsencrypt by swag (Dec 30th 2021)

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!