Improvements: Run web only if enabled + Rename svc-beets + Add s6 overlay for drop2beets

[dtrunk90]

---

• I have read the contributing guideline and understand that I have made the correct modifications

---

Description:

1. Run web only if enabled: This fixes [FEAT] Running the docker container without starting up a web instance #141
2. Rename svc-beets: In order to reflect what this s6 overlay does it's better to name it beets-web since the only purpose of this overlay is to run the web plugin as a longterm service
3. Add s6 overlay for drop2beets: This is not enabled by default but the required s6 overlay to be able to enable it was added here

Benefits of this PR and context:

It fixes #141 and all users would benefit from the other commits

How Has This Been Tested?

I've used the following docker-compose.yml to run and test my changes:

services:
  beets:
    build: .
    container_name: beets
    environment:
      - PUID=1000
      - PGID=1000
      - TZ=Europe/Berlin
      - DOCKER_MODS=linuxserver/mods:universal-package-install
      - INSTALL_PIP_PACKAGES=drop2beets
    volumes:
      - ./config:/config
      - ./dropbox:/dropbox
      - ./logs:/logs
      - ./music:/music
    restart: unless-stopped

• I've checked w/o plugin web if the error message error: unknown command 'web' dissappeared.
• I've checked w/ plugin web if the web service is up and running
• I've checked w/ plugin drop2beets if the service is up and running

Source / References:

#141
#138

[github-actions]

Thanks for opening this pull request! Be sure to follow the pull request template!

[LinuxServer-CI]

I am a bot, here are the test results for this PR:
https://ci-tests.linuxserver.io/lspipepr/beets/2.5.1-pkg-1185004a-dev-f1f827b03fd0ed097b96b88a3e05c22f0bb5a59c-pr-142/index.html
https://ci-tests.linuxserver.io/lspipepr/beets/2.5.1-pkg-1185004a-dev-f1f827b03fd0ed097b96b88a3e05c22f0bb5a59c-pr-142/shellcheck-result.xml

Tag	Passed
amd64-2.5.1-pkg-1185004a-dev-f1f827b03fd0ed097b96b88a3e05c22f0bb5a59c-pr-142	✅
arm64v8-2.5.1-pkg-1185004a-dev-f1f827b03fd0ed097b96b88a3e05c22f0bb5a59c-pr-142	✅

[LinuxServer-CI]

I am a bot, here are the test results for this PR:
https://ci-tests.linuxserver.io/lspipepr/beets/2.5.1-pkg-1185004a-dev-ebffeef570b82696f320b9b7fb1ae10e973a76fd-pr-142/index.html
https://ci-tests.linuxserver.io/lspipepr/beets/2.5.1-pkg-1185004a-dev-ebffeef570b82696f320b9b7fb1ae10e973a76fd-pr-142/shellcheck-result.xml

Tag	Passed
amd64-2.5.1-pkg-1185004a-dev-ebffeef570b82696f320b9b7fb1ae10e973a76fd-pr-142	✅
arm64v8-2.5.1-pkg-1185004a-dev-ebffeef570b82696f320b9b7fb1ae10e973a76fd-pr-142	✅

[LinuxServer-CI]

I am a bot, here are the test results for this PR:
https://ci-tests.linuxserver.io/lspipepr/beets/2.5.1-pkg-1185004a-dev-f5cfc0cb136d2bbd3c9c02dc7cb4d0c93161d339-pr-142/index.html
https://ci-tests.linuxserver.io/lspipepr/beets/2.5.1-pkg-1185004a-dev-f5cfc0cb136d2bbd3c9c02dc7cb4d0c93161d339-pr-142/shellcheck-result.xml

Tag	Passed
amd64-2.5.1-pkg-1185004a-dev-f5cfc0cb136d2bbd3c9c02dc7cb4d0c93161d339-pr-142	✅
arm64v8-2.5.1-pkg-1185004a-dev-f5cfc0cb136d2bbd3c9c02dc7cb4d0c93161d339-pr-142	✅

[LinuxServer-CI]

I am a bot, here are the test results for this PR:
https://ci-tests.linuxserver.io/lspipepr/beets/2.5.1-pkg-87cc8640-dev-b1dca77ba4a65a1c0a56c4b73f7cf0560267ae24-pr-142/index.html
https://ci-tests.linuxserver.io/lspipepr/beets/2.5.1-pkg-87cc8640-dev-b1dca77ba4a65a1c0a56c4b73f7cf0560267ae24-pr-142/shellcheck-result.xml

Tag	Passed
amd64-2.5.1-pkg-87cc8640-dev-b1dca77ba4a65a1c0a56c4b73f7cf0560267ae24-pr-142	✅
arm64v8-2.5.1-pkg-87cc8640-dev-b1dca77ba4a65a1c0a56c4b73f7cf0560267ae24-pr-142	✅

[dtrunk90]

Would be great to get at least approval for the first commit which conditionally runs web (or drop that service entirely). The beets custom service could be done via a custom service volume.

[aptalca]

My thoughts:

• drop2beets should be a mod, see https://mods.linuxserver.io/
• web is the only longrun service that runs in this container. If you remove that, the container becomes a oneshot container, which is not how it was designed or intended. Plus, I don't see the point of removing the web service as its resource usage is negligible and you can always not map a port and ignore it completely.
• [moot point] the proposed if blocks in the services will have the script getting restarted by s6 if the plugins are disabled as the script will keep exiting.

[dtrunk90]

* drop2beets should be a mod, see https://mods.linuxserver.io/

It is. The only thing done here was adding an s6 service. But as mentioned in my previous comment this could also be achieved by mapping a volume.

* web is the only longrun service that runs in this container. If you remove that, the container becomes a oneshot container, which is not how it was designed or intended. Plus, I don't see the point of removing the web service as its resource usage is negligible and you can always not map a port and ignore it completely.

Agree and disagree. Not all users want to use the container as a web service. Some of us want a drop2beets container, which is a longrun service too. I understand your point but why making a difference for drop2beets then? Either treat it equally (add drop2beets service and let the users choose what longrun service they actually want by checking their config like i did for both of the s6 services and add the s6 service for drop2beets as well or drop the web s6 service as well since it could be a mod, right?

* [moot point] the proposed if blocks in the services will keep getting restarted by s6 if the plugins are disabled as the script will keep exiting.

Is that how s6 works? It retries over and over again?

[aptalca]

It is. The only thing done here was adding an s6 service. But as mentioned in my previous comment this could also be achieved by mapping a volume.

That's not a mod (referring to our docker mods as linked). Mods are self contained tarballs that inject everything they need runtime (init and service files, binaries, package install lists, etc.) and need no other modifications to the base image prior.

Agree and disagree. Not all users want to use the container as a web service. Some of us want a drop2beets container, which is a longrun service too. I understand your point but why making a difference for drop2beets then? Either treat it equally (add drop2beets service and let the users choose what longrun service they actually want by checking their config like i did for both of the s6 services and add the s6 service for drop2beets as well or drop the web s6 service as well since it could be a mod, right?

We designed a beets container with the web service. If you want something else, feel free to fork it and build it yourself, or you can mod it via different ways: https://docs.linuxserver.io/general/container-customization/

Is that how s6 works? It retries over and over again?

That's its job as the supervisor. If a service stops, it restarts it.

[dtrunk90]

It is. The only thing done here was adding an s6 service. But as mentioned in my previous comment this could also be achieved by mapping a volume.

That's not a mod (referring to our docker mods as linked). Mods are self contained tarballs that inject everything they need runtime (init and service files, binaries, package install lists, etc.) and need no other modifications to the base image prior.

True. That's just customization - not what can be considered a mod.

Agree and disagree. Not all users want to use the container as a web service. Some of us want a drop2beets container, which is a longrun service too. I understand your point but why making a difference for drop2beets then? Either treat it equally (add drop2beets service and let the users choose what longrun service they actually want by checking their config like i did for both of the s6 services and add the s6 service for drop2beets as well or drop the web s6 service as well since it could be a mod, right?

We designed a beets container with the web service. If you want something else, feel free to fork it and build it yourself, or you can mod it via different ways: https://docs.linuxserver.io/general/container-customization/

Tbh that's too much work just to remove the web s6 service. I'm not really happy with that solution but I do understand the point here.

Is that how s6 works? It retries over and over again?

That's its job as the supervisor. If a service stops, it restarts it.

I just thought it wouldn't if the exit code is 0 but thanks. Learned something new.

I'll think about submitting a mod then.