My observation towards OMV-5 : Docker infestation, Improper docker use.

  • Hello, Even though I am a newbie in this community and just trying to get familiar with OMV environment, using OMV 5, I would like to share my two cents relevant to use of docker with this community.


    I have learned that after posting several questions to the forum, many plugins will not be available in OMV-5. One of the responses stated that

    You're far better off embracing docker at this point, since that is going to be where almost everything is going. Once you get the hang of it, it is incredibly simple.


    I understand importance of Docker, benefits. But the way that docker is being planned in OMV-5, I suspect that is beneficial for the future of OMV.


    My response to the above


    Honestly, even if docker will be used, it should fit into the system conventions. For example, I am advised to use a container in place of webdav plugin in a new thread, because webdav will not be supported in OMV5. I installed it, now I have fiddle with new users, permissions etc. in docker environment. I have to make docker volumes attach to the shares and many other tids and bits.


    This corrupts whole nas idea. About, more than ten years ago, I was managing my personal linux servers with samba, nfs, ftp etc. I was tired of these, managing users, shares, permissions in different environments and applications and bought a nas. NAS is supposed to facilitate and enhance. Spinning docker containers from all sources is not solution to the problem.


    If I am to manage docker containers for even basic NAS services, I can just get rid of OMV, make every service docker container and just run my bare linux server with docker containers. Why will I be needing OMV or any other NAS solution.


    I want to use docker to run applications of my interest. However many services are dependent on OMV settings, configuration such as users, shares etc. If OMV will quit plugin architecture, and move to docker, I believe Docker containers must be build as a way that it will integrate with OMV. Users should not be forced to hassle with configuration of containers to bridge with OMV


    If OMV is being left to provide a couple of basic services, and all others are left to docker containers from any sources, then surely in a very short amount of time, OMV itself will be obsolete since it will be bringing very little to the table.


    Whole NAS idea was to minimize these management activities and provide network storage, later people wanted to utilize NAS devices with applications, which I also want. But this should not hamper the fundamental NAS idea and ease of use.


    My proposal is simply if plugins will be container, OMV should provide containers that fit the infrastructer seamlessly and with little or no effort. OMV may itself be a docker container or can be run as a stack (I do not know if this is possible in terms of managing hardware). But in anycase OMV should provide eco-system of containers that fit into the OMV environment seamlessly if docker will be substitute of OMV plugins.


    I may want to run none OMV certified/released container than I have to do the chores by myself. Other than that, these plugin substitute containers must be build by the plugin developers for the best use of OMV.


    Current approach seems beneficial to developers NOT to the end users. I am, as an end user, here for ease of use, sparing time from these chores not to add new ones.


    These are my personal options, I hope I would not offend anybody.


    Thanks.

  • Hello,


    I am completly new to OMV (moved from debian 9.0 server). My idea was to move to a proper nas solution because really my debian server was an overkill considering I was mostly using plex and file sharing. I am not a Linux expert so I was looking for something simple.


    I am very pleased with the web console admin interface, the ease of creating RAID and Samba share ... but I am very suprised of the lack of support of plex. I believe most of the user who will look for distrib like OMV will expect out of the box plex support.


    Maybe developer can think of a way to make it transparent to user, something like a scipt that will install docker, create the docker user, download the plex image, create and map the correct directory .... I don't know if this is feasible but for sure that would be nice !


    Otherwise it might be a stupid question, but instead of a docker image and considering the base is Debian, would it be possible to simple do an apt-get install mediaplexserver ?

  • You are right, containers are not the final solution. But the problem for the current developers is that there are too much plugins and less developers. The eco system of fully integrated (non Docker) plugins can only increase if there are more developers.


    The current situation is that 99,9999999% users are only claiming for plugins, but none of them are willing to contribute to the project. I think you see the problem. It is like everywhere, nowadays people think everything is for less, but nobody thinks about that such a project like OMV or other OSS projects can only grow and expand with the help of others; and that is work, so nothing happens.

  • Also the configuration backend has changed, more work for porting. The webui is gonna change, more work for porting, nobody now is the php backend is going to change, could more work in the future for porting or developing something new


    But you're correct there should me more integration for sharing services, like you said webdav or could be minio or another service as a plugin. But omv needs to settle is codebase, until then nothing new will come in terms of plugins. OMV accomplish the basics in LAN with SMB/NFS, FTP and Rsync in a very simple and effective way.


    Running my server now in proxmox i took off docker from omv, and now i run exclusively a VM that hosts all dockers

  • @subzero79 The next big thing is the new UI in OMV6 which is sadly absolutely necessary because Sencha does not ship any ExtJS updates anymore and the latest ExtJS framework is not GPL'ed anymore. So i have to switch to a open framework which is Angular and Angular Material in this case. The planned REST API for OMV7 is optional, the PHP RPC interface will still exist in parallel. So with OMV6 the API and UI should be a stable foundation.

  • many plugins will not be available in OMV-5

    Most of the plugins (downloader plugins) that aren't available on OMV 5 weren't available on OMV 4. And these plugins belong in docker due to the bleeding edge requirements and security issues. There are still plenty of plugins available that will never be docker - https://forum.openmediavault.o…?postID=186122#post186122


    If OMV will quit plugin architecture, and move to docker

    I don't know why you think this will happen. Some things cannot move to docker.


    If OMV is being left to provide a couple of basic services, and all others are left to docker containers from any sources, then surely in a very short amount of time, OMV itself will be obsolete since it will be bringing very little to the table.

    Your fear of/lack of experience with docker is what makes you think this. There is a reason containers are used and we are trying to move the plugins that belong in containers to containers. Not just blindly moving things to containers.


    Whole NAS idea was to minimize these management activities and provide network storage, later people wanted to utilize NAS devices with applications, which I also want. But this should not hamper the fundamental NAS idea and ease of use.

    Which plugin that provides "normal" NAS applications is moving to docker? I'm sorry that I disagree that webdav is not a normal NAS application even if the commercial NASes provide it. They provide lots of non-NAS applications now.


    Current approach seems beneficial to developers NOT to the end users. I am, as an end user, here for ease of use, sparing time from these chores not to add new ones.

    That is great but the end users are not helping me. There aren't developers. Really just me and some occasional help. I have to do what I have time for. If this does not fit, maybe another product would be a better solution.


    These are my personal options, I hope I would not offend anybody.

    Not offending but also not realistic. As Volker said, until there is more help, none of this will change.

    omv 5.5.17-3 usul | 64 bit | 5.4 proxmox kernel | omvextrasorg 5.4.2
    omv-extras.org plugins source code and issue tracker - github


    Please read this before posting a question.
    Please don't PM for support... Too many PMs!

    Edited once, last by ryecoaaron ().

  • but I am very suprised of the lack of support of plex. I believe most of the user who will look for distrib like OMV will expect out of the box plex support.

    I struggled supporting the plex plugin for years (i hate plex too). It was terrible and doesn't really make sense since the plugin just enabled the service and gave a link to the plex web interface. No matter how you have plex installed or how advanced the plugin could be, using the plex web interface is still required. So, why have a plugin to enable when a docker does the same?

    omv 5.5.17-3 usul | 64 bit | 5.4 proxmox kernel | omvextrasorg 5.4.2
    omv-extras.org plugins source code and issue tracker - github


    Please read this before posting a question.
    Please don't PM for support... Too many PMs!

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!