Linuxserver.io Nextcloud not updated?

  • I have very nice installation of LE with docker add-ons from the LE repository, that wrapp the docker images form Linuxserver.io. I have SQL-DB, Portainer and Nextcloud. The Updater is also installed and from time to time I see the message that new docker images are installed. BUT: when I check on Linuxserver.io the Nextcloud image is already on V.17.x while my Nectcloud is still V.14.x. What is going wrong or do I need to install/update something manually?


    Any help or experience somebody?

    /BJ

  • Nextcloud is one of the few images we provide that require in app updates.


    As long as the docker image is up to date (auto updated and applied on reboot), you can do the gui updates to update nextcloud

  • ok

    1. I see that I can update within the nextcloud webgui. I will do so.

    2. How can I be sure, that I have the newest image from Linuxserver.io loaded and active ? I checked in Portainer and it shows that in "Container" that it was created (locally) today and in "Images" that is was built at BUILD_DATE=2019-12-26T19:44:02+01:00 . That seems to be proof enough?

    3. While checking the images at .../.kodi/userdata/addon_data/service.system.docker/docker/.... I found a lot of old (orphaned?) images/overlays. But I read that the updater has now a prune-function. When I check in the LE gui and open at the updater add-on the configuration page I see no entries on a blank screen with a crippled header?? What goes wrong here, what to do ?


    /BJ

  • The image label you're seeing in portainer suggests you are using the latest image.


    The settings screen looks strange. What hardware platform are you on? I'll do some testing here.


    In the meantime, you should be able to check for old images via portainer or cli.

    1. I am using LE Generic on Intel, it looks like the broken form for updater config is due to some theme incompatibility ?
    2. I have upgraded successfully from 14 to 15 to16 to 17, needed some retries and nextcloud "occ" commands to correct some indices.
    3. I would still like to understand, how can I update from nextcloud V14 to V17 while the docker image is always pulled from Linuxserver.IO. Is that still on V14 ?? and what happens to my updates when a new image is pulled ??

    Nevertheless, many thanks for your work and support, it is really appreciated !!

    /BJ

    1. I am using LE Generic on Intel, it looks like the broken form for updater config is due to some theme incompatibility ?
    2. I have upgraded successfully from 14 to 15 to16 to 17, needed some retries and nextcloud "occ" commands to correct some indices.
    3. I would still like to understand, how can I update from nextcloud V14 to V17 while the docker image is always pulled from Linuxserver.IO. Is that still on V14 ?? and what happens to my updates when a new image is pulled ??

    Nevertheless, many thanks for your work and support, it is really appreciated !!

    /BJ

    For nextcloud, all application files are stored in the bind mounted folder "/config". That's why updating the docker image doesn't update your nextcloud install and that's why the update won't be reverted.


    Essentially, our images come with a set of nextcloud app files but they are only installed if it's a brand new install. If not, the existing install is untouched.


    We had to do it that way because nextcloud often has to perform various tasks in between updates and we have no control over those. This way we let the web (or occ) updater perform those tasks including various database updates and modifications.


    I would also recommend deleting the file at /config/nginx/site-confs/default and restarting because it doesn't automatically update (user customizable file) and there were some necessary changes for the latest version. Otherwise you'll get some warnings about headers in the admin settings

  • ok,

    that explains a lot (I found the old images in /config and was surprised!).

    And deleting /config/nginx/site-confs/default and the following update will not touch my data ?


    I did it and after restart I see a new "default" file but the log shows many lines with

    "nginx: [emerg] duplicate upstream "php-handler" in /config/nginx/site-confs/default:1"

    and the web-guy is not reachable.....?

    I try to force a pull now.


    Sorry: was my fault, I kept the old "default" by just renaming and it took both files then. I deleted the old and it is now up and running !!


    Thanks,

    /BJ

  • Any idea about the crippled configuration page of the updater add-on in LE ?


    Thanks,

    /BJ

    Finally figured out the fix for the elusive bug. It seems it was a regression in kodi. But it's fixed now.

  • Back from a trip I see your message, I tested and it looks like the add-on updated to V.1.04

    But I open the configuration page and see the same crippled picture as before!

    Maybe a fresh install would do it, but this is not possible because the other linuxseerver.io addons require it and so can not be deactivated/deleted etc....


    Any idea ?

  • Back from a trip I see your message, I tested and it looks like the add-on updated to V.1.04

    But I open the configuration page and see the same crippled picture as before!

    Maybe a fresh install would do it, but this is not possible because the other linuxseerver.io addons require it and so can not be deactivated/deleted etc....


    Any idea ?

    In your userdata folder for the updater addon, there should be an empty settings.xml file. Delete that and restart the device. With the new version, that file won't be created anymore (which was the original issue where kodi leia somehow fails to read the addon's main settings file if there is a blank but properly formatted user xml in userdata)