Portainer: Much Ado About Nothing

  • I would agree, however, Portainer is quite straightforward and the use of docker compose files using stacks is an easier way to deploy a container, have a look in the guides section for some examples, and each container in docker hub will usually have a compose file which requires editing to point to your own host shares.

    100% agree. I still find Portainer really easy to set up containers, just by looking at docker-compose files, or looking at docker-run commands. Yeah some of that is learning where to look for things, etc.. but that's why I spent my last few weeks on OMV 4 learning Portainer, before moving to OMV 5.

  • I still find Portainer really easy to set up containers,

    TBH it just helped me resolve a problem with CloudCMD, the latest version wasn't displaying correctly, looking at github there were other users with the same problem and the solution was to use a previous version.

    By using stacks it makes it easier to redeploy the container with an older but working version, you have to do some housekeeping afterwards but that is a small price to pay.

    But because of the above problem I have stopped Watchtower as it's just as easy to update manually when necessary.

  • using docker compose still required me to to do a lot of writing and copy pasting. e.g. now to map a directory to my docker container i need to browse to it trough ssh and copy&paste its paath, as there is no option to browse to it via the gui. votdev usually used to claim the OMV is supposed to be configured trough the gui, but now a lot of things are being moved to the command line.

    HP Proliant Microserver Gen8, G2020T, 4 GB ECC RAM, 3x WD RED 3TB + 1x WD RED 4TB + OCZ Vertex3 128 GB SSD

  • usually used to claim the OMV is supposed to be configured trough the gui, but now a lot of things are being moved to the command line.

    Everything Volker has done is configured through the GUI. For me, portainer was a better option because it is well supported and has many more options than the docker-gui plugin. But even with that plugin, it didn't support docker-compose. You can use docker-compose files with Portainer Stacks keeping you from using the scary command line.

    container i need to browse to it trough ssh and copy&paste its paath, as there is no option to browse to it via the gui

    Use the symlinks plugin. Then you can use the magical file explorer in the omv web interface to make a symlink to the location you need.

    omv 5.5.1 usul | 64 bit | 5.4 proxmox kernel | omvextrasorg 5.3.3
    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!

  • its not about the command line being scary. its about either using the command line or a gui and not swithing between them. and portainer being a UX nightmare.

    pulling an image? you can't do that from portainer as there is no search function in portainer itself, need to go to dockerhub, copy its name and insert it into portainer,

    crearting a container? there's no apparent way to create a container and in the container section, you need to go to containers, but there you cant see your unused images. so again, back to images, check its name, copy back to containers and paste

    setting up container paramaters? you need to copy everything from the dockerhub readme, as portainer is not showing the preset variables like docker-gui used to, or cockpit does (which has it's own serious flaws as well)


    setting up a container in docker-gui was a few clicks, now i need to write a novel to do it

    HP Proliant Microserver Gen8, G2020T, 4 GB ECC RAM, 3x WD RED 3TB + 1x WD RED 4TB + OCZ Vertex3 128 GB SSD

  • i thought id chime in as i posted in the earlier days of the thread,


    i decided to update my system to omv5 but did a fresh install tried to clonezilla first so i had a backup but something was wrong with the ssd it couldn't do it wouldn't install a fresh version either had to swap it out. however i backuped up my /config folders from where they was stored so move to there new homes in a appdata share on the raid, and took this time to learn compose files just watched a single video of technodad who did a portainer unifi controller compose set up (needed to do this container anyways knowing this i then managed to do portainer stacks for all my other containers as well didn't take long at all.

    OMV 3.0.58 - 64 bit - Nut, SABnzbd, Sonarr, Couchpotato
    HP N40L Microserver, 8gb Ram, 5 x 3TB HDD Raid5, 1 x 120GB 2.5" SSD (OS)

  • now i need to write a novel to do it

    I started this thread because I thought that Portainer could be daunting for new users, but TBH it's not and some longstanding members have created compose files and added them to the Guides section.

    Portainer has 155 contributors, OMV on the other hand has ryecoaaron who maintains the plugins, can you see the difference. The likes of OMV, Freenas, Xigmanas have WebUI's which assist the end user to configure and maintain what is in effect a 'headless' machine, no Desktop. So anything written and developed for these systems is a bonus as it helps us the end user.

    This guy on youtube has done a number of videos on docker and OMV5, he actually uses compose files and stacks but he locates the share path by using googles inspect.

    You don't have to re write War and Peace, it's just adapting to a new way of doing things, and searching on docker hub is a bonus how do you think TechnoDadLife found his images for his videos.

  • using docker compose still required me to to do a lot of writing and copy pasting. e.g. now to map a directory to my docker container i need to browse to it trough ssh and copy&paste its paath, as there is no option to browse to it via the gui. votdev usually used to claim the OMV is supposed to be configured trough the gui, but now a lot of things are being moved to the command line.

    If you're using Portainer, literally all you have to do is copy/paste the docker-compose file that is on many containers on docker-hub, into a stack, name the stack.. make adjustments (paths, puid, etc.) just as you would have with the docker-gui plugin.. and hit deploy.


    How is this requiring command line or writing a novel?

  • its about either using the command line or a gui and not swithing between them. and portainer being a UX nightmare.

    OMV is linux. It is not uncommon to use the command line. I realize that is weird in the Windows world but there is nothing wrong with it. Hell even Windows is moving to more powershell stuff that is more like Linux. I don't mind portainer's web interface myself. I only use docker from the command line but I can find everything I need easily in portainer.

    pulling an image? you can't do that from portainer as there is no search function in portainer itself, need to go to dockerhub, copy its name and insert it into portainer,

    Being able to search in the plugin and just blindly selecting the first (or second or third) choice without reading about the image is a bad idea. You *should* be reading about the image first on docker hub before pulling it.


    setting up container paramaters? you need to copy everything from the dockerhub readme, as portainer is not showing the preset variables like docker-gui used to, or cockpit does (which has it's own serious flaws as well)


    setting up a container in docker-gui was a few clicks, now i need to write a novel to do it

    More options and flexibility means more clicks. Lots of people complain that OMV itself is too many clicks. The plugin isn't coming back unless someone else maintains it. So, not sure what to tell you.

    omv 5.5.1 usul | 64 bit | 5.4 proxmox kernel | omvextrasorg 5.3.3
    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!

  • It’s human nature to resist change. Nobody likes throwing away old “perfectly good” tools. And there’s the learning curve to overcome - again. First there was the fuss that containers were replacing many plugins. When that pretty much died down, the resistance to Portainer began. I dislike change too. However change can be good.


    molnart I am not in the same class as these above who have been responding to you, so I probably should just leave it to them. On the other hand, maybe an encouraging word from someone more on your level will help.

    • This @TechnoDadLife video - which translates from Docker to Portainer was a big help to me.
    • My OMV journey began with the Docker GUI and I can remember saving a container in Docker and getting a folder path wrong (no leading slash), or forgetting to click that button on the far right of a variable line. Then I would have to delete the container and start over from scratch. In Portainer that never happens. If you forget a leading slash in a path or some such you are not allowed to save the container. You have to go back and correct your mistake.
    • For most (All?) containers you can ignore most of the sections of Portainer. Just concentrate on the sections you used in Docker GUI and find out where they are in Portainer. That’s where the TechnoDadLife video above is so helpful.
    • If you don’t have an extra machine - get one (if you can) and set up an OMV testing server. I once had Plex container running on four machines at once just testing things. Any machine will do. A inexpensive RPi4, an Odroid HC2 or a dual core intel desktop someone is about to throw out. Just about anything will work.
    • Stacks in Portainer is really quite easy. Go to https://hub.docker.com/r/linuxserver/plex and copy the docker-compose file and paste it into Portainer/Stacks and insert your personal variables and paths (you did the same thing when you were setting up a container in Docker GUI). It just works.
    • For that matter, once you get over command line phobia docker-compose on the command line is even easier.
    • If you get stuck there is plenty of help on the forum - in archives, and in live help from guys like these above (who are really trying to help you now). If it’s not too deep I might even be able to help.
    • If I remember correctly when I first started with OMV the command line was required almost immediately. All of the TechnoDadLife videos I ever used had some command line work in it. Yes, it was scary at first but I got over it. You will too.
    • I don’t know what else to say. Go back over the posts above and re-read what these guys are trying to get across to you. They really want you to succeed. I wish you well too.

    RAID is NOT a backup and not useful for most home users. Rsync makes true backup and restoration stupid easy, and it's built right in to OMV. Use this command in a Scheduled Job: rsync -av --delete /srv/dev-disk-by-label-NAMEofSOURCEdisk/ /srv/dev-disk-by-label-NAMEofDESTINATIONdisk/

    Hardware: OMV 5 (current) - NanoPi M4: Nextcloud, Plex, & Heimdall - Acer Aspire T180: backup - Odroid XU4: Pi-Hole (DietPi) - Odroid HC2, Raspberry Pi 3B+, and HP dx2400: testing.

  • Portainer has 155 contributors, OMV on the other hand has ryecoaaron who maintains the plugins, can you see the difference.

    It seems he alone did a much better job then the 155 Portnianer developers.


    OMV is linux. It is not uncommon to use the command line.

    I have no issues with the command line. For few years i have used archlinux as my daily OS, deployed from scratch in command line. Linux console is probably my most used app in windows. my issue is elsewhere.



    You *should* be reading about the image first on docker hub before pulling it.

    Yes. Therefore Portainer should pull this page from dockerhub and show it in its interface. Container deployment is a fairly simple job. Requiring for that to navigate multiple pages in a single application is an utter failure. Requiring to open websites and ssh terminals for a simple job is deliberate annoying of the users. Again I am used to looking for solutions and googling them in multiple sources, running troubleshooting via terminal etc. but that's troubleshooting, not a simple standardized activity.


    he plugin isn't coming back unless someone else maintains it. So, not sure what to tell you.

    I get it. I just need to ventilate my hate on Portainer and nostalgize over docker gui, which of course was far from ideal but in terms of ux significantly better than portainer.


    If you don’t have an extra machine - get one (if you can) and set up an OMV testing server.

    That's what virtual machines are for.

    Stacks in Portainer is really quite easy. Go to https://hub.docker.com/r/linuxserver/plex and copy the docker-compose file and paste it into Portainer/Stacks and insert your personal variables and paths (you did the same thing when you were setting up a container in Docker GUI). It just works.

    Except when you use macvlan for most of your containers (e.g. all of them besides Plex). I haven't found yet a simply understandable description how to set this up via docker compose. Read: setting this up in portainer was easier for me for my 7 running containers than finding out how to do it via docker-compose.

    HP Proliant Microserver Gen8, G2020T, 4 GB ECC RAM, 3x WD RED 3TB + 1x WD RED 4TB + OCZ Vertex3 128 GB SSD

  • Therefore Portainer should pull this page from dockerhub and show it in its interface.

    Then make a feature request on github :)

    I just need to ventilate my hate on Portainer and nostalgize over docker gui

    Hm, then the option would be to stay on OMV4

    Requiring for that to navigate multiple pages in a single application is an utter failure.

    No, you don't have to navigate multiple pages, using stacks just requires the .yml file to be edited to point to your host's share.


    But I get it you don't like it, but it's here to stay

  • [...]

    Hm, then the option would be to stay on OMV4

    Looks like I will just do that for my RPi installations for the foreseeable future. I recently tried to migrate an OMV4 installation to OMV5 (using #38) and that didn't work as expected. Luckily it was a working clone of a VM. And just for experimenting, I already have two OMV5 VMs (one more or less standard, the other with Proxmox). I can find my way around in Portainer, but just kicking off an image in the old Docker GUI is still a lot easier, all settings for a container are on one page instead of switching between tabs with the tab buttons somewhere half way down the page instead of near the top where they can stick in a fixed position.

  • but just kicking off an image in the old Docker GUI is still a lot easier

    I agree, the search function in the Docker GUI is seriously lacking in Portainer.

    but just kicking off an image in the old Docker GUI is still a lot easier, all settings for a container are on one page instead of switching between tabs with the tab buttons

    Again I agree, but using Stacks and editing the Stack to point to the hosts shares is a lot easier than the plugin. I've had a problem with CloudCMD, this was deployed through the plugin, but to resolve and test, the stack option is easier all I do is change the containers 'tag' in the Stack then update. A new image is pulled and deployed.

  • Except when you use macvlan for most of your containers (e.g. all of them besides Plex). I haven't found yet a simply understandable description how to set this up via docker compose. Read: setting this up in portainer was easier for me for my 7 running containers than finding out how to do it via docker-compose.

    You do it the same way you do it in the plugin - you create the macvlan network first in the network section (portainer has documentation - https://www.portainer.io/2018/09/using-macvlan-portainer-io/) and then you specify it in the container or docker-compose/Stacks section.

    omv 5.5.1 usul | 64 bit | 5.4 proxmox kernel | omvextrasorg 5.3.3
    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!