primeval_god

These are not exactly in the unRAID filesystem. The 4 above use volume syntax so docker volumes were created to back those mount points. They are named teslamate-db, teslamate-grafana-data etc. You should be able to see them using the command docker volume ls You would also need to use docker volume commands to delete them or copy data out of them.

Personally i cant say that I am a fan of more "modern" UIs like HexOS. I have always quite liked unRAID's Azure and Grey theme layouts, and thought they would adapt quite well to mobile assuming (probably naively) that the code could be made to scale correctly.

You must specify a folder that does not exist and let the plugin create the folder.

The script should not be necessary. The error message is because you specified "/mnt/cache" as the location for the swapfile on a BTRFS drive. As the help text for the swapfile location notes you need to choose a non-existing directory name. The plugin will then create a subvolume at that location with the proper options for the swapfile.

Looks to me like the first screenshot is of the APPs page not the compose page. Is there a problem with APPS as well?

If you use a BTRFS or ZFS filesystem, I would highly recommend taking periodic read only snapshots of your data. Especially if you dont have a lot of churn on your files the space cost for CoW snapshots is negligible and having them as a safety net helps the peace of mind.

I think this is part of the issue https://github.com/bergware/dynamix/issues/94

Unraid is based on Slackware, which is not (so far as i can see) a distro supported by sysbox.

Templates available in CA are not necessarily maintained by the creators of the applications or docker images they provide. In the case that the template and image are maintained by the application creators i believe there is a specific "Official" tag applied by CA. For the official policies of what app templates are included and how they are moderated see this thread

You should just be able to click the up button. Bring the stack completely down and then back up shouldn't be necessary, but it should definitely cause your stack to pickup any changed values.

I will consider it, however such a feature runs counter to the general direction of the plugin. My goal for the plugin is to have a minimal set of features to support simple single file compose stacks.

Apparently i didnt read the info closely enough. It looked to me like they were showing referencing multiple env files by absolute path from within a compose file.

As far as I am aware there are no limitations that would prevent you from doing this. There is no support for editing the extra env files in the compose manager ui however.

LanCache might work for this.

You can run compose commands from the terminal as compose is installed system wide. You will however need to add the -f flags to specify the paths to your compose files.

This (along with the wording in the full announcement) seems to suggest that it will continue to use a plugin architecture probably like the Unraid Connect plugin.

Here is a link to the bug report.

Yes this is pretty much expected. Unraid is meant to boot from a usb flash drive. An SD card is not a flash drive. Some people have found sd card readers that will work, but in general it's not expected that they will. When it comes to booting from flash cards it's not about the card it's about the reader. The reader, which is the usb device, must present a unique GUID which most do not.

Does the problem recur when switching back to compose in the docker tab?

@bmartino1 I cant really tell from your post what the issue with your grafana stack actually is. Is it something to do with the webui link or a continually restarting container? This plugin is not currently tested against unraid v7 (by me at least) so I am not yet familiar with the docker system changes for the beta3 version. If the issue is with the container rather than the webui integration could you tell me what version of the docker engine unraid v7 has?

I am not aware of any changes that would cause the issue, nor have I seen any other reports of it. To debug I would suggest opening the developer options in your browser and watching the console for errors when reloading the docker page. Additionally in the latest version of this plugin there is an option (in the settings page) to break the compose plugin out of the docker page into it's own top level tab. You could try enabling that option and see if it's the compose tab of the docker tab that ends up being non responsive.

I am not a python expert but I dont think that warning is particularly relevant in a single application docker container. There is not much system in the container to worry about interfering with. As for the second part I dont think a venv would help, unless its base directory was stored outside the container. What you need is a way to run the install commands only on the first startup of the container. Some containers have a framework for some sort of first run script. You could achieve the same by checking for and creating a lock files somewhere in the containers internal filesystem.

Did you create the container using the unraid dockerman interface or via some other method?

A custom docker network? You can specify any existing docker network in your run command. I cant think of one off hand. For a python script container I cant think of much unRAID specific stuff to worry about.

I am guessing that ssd1 is a disk mounted with unassigned devices? The plugin wasnt really designed with those in mind. A permanent single disk pool is really more suited to hosting a swapfile.