Additional features

[LiquidLife86]

Add the ability to scale the instances
Add the ability to choose what image is used for each service

Give fare warning you have referral codes baked into the code for each of theses.

[GeiserX]

1. ok
2. How do you see the flow? I'll brainstorm with AI
3. I added it but after the rework in the docs it was gone from README.md somehow (EDIT - I now realized it was still there), reinstating, and adding it as well on the signup links inside the app

thanks @LiquidLife86

[LiquidLife86]

When you go to the service catalog, you only see the services available (earn app, earn.fm etc) there isn't a way to choose the docker image used.

For instance I could not get the earnapp to show an online device so I rewrote the earnapp.yaml to use the venatum/earnapp image instead

[LiquidLife86]

Well, I deployed it using the advanced and using the venatum/earnapp:latest but it keeps loop restarting.

[LiquidLife86]

Gordon (Docker Desktop AI) says:

It keeps restarting because /sbin/init exits with code 255, and the container’s unless-stopped restart policy starts it again. The logs are empty, so the exact failure isn’t shown; the configured ALL capability drop and no-new-privileges setting may be preventing the image’s init process from starting. The other earnapp container’s exit code 137 is a separate event and doesn’t establish the cause here. Let me know if you have any questions!

[GeiserX]

That one is on me: I told you to put venatum/earnapp in the box, and without a tag that is latest, which is the systemd image. Its main process is /sbin/init, and systemd needs a privileged container with the host's cgroup tree. CashPilot never grants that (every deployed container drops all capabilities), so it exits 255 immediately, logs nothing, and the restart policy brings it back forever. Gordon read it right.

Use venatum/earnapp:lite instead. Same EARNAPP_UUID, no systemd. I ran both lite images on one of my boxes under exactly the hardening CashPilot applies and both register and report enabled; venatum/earnapp:latest exits 255 every time.

Two things changed so nobody else walks into this: the worker now refuses an image that boots an init system at deploy time, with the reason in the error and without touching the running container, and the EarnApp guide names venatum/earnapp:lite as the alternative to try when the device never shows online. It will be in the next release.
#453

Once you are on :lite, does the device show online in your EarnApp dashboard? If venatum's lite works where fazalfarhan01's does not for you, I'll look at the default.

[LiquidLife86]

Tried with venatum/earnapp:lite but it did the constant restarting still.

Tried again with the default fazalfarhan01/earnapp:lite and its good now and earnapp shows a device online.

Heck ya man! GOOD job!

[GeiserX]

Glad the default works. I couldn't reproduce the venatum lite loop: on my box it stays up under the same limits CashPilot uses. One catch is that its script restarts EarnApp inside the container and logs "restarting in 5s", which looks like a loop even when the container itself is fine. If you have a minute, docker inspect <container> --format '{{.RestartCount}} {{.State.ExitCode}} {{.State.OOMKilled}}' on the venatum one would tell us which it was. Either way the default stays fazalfarhan01's lite.

[LiquidLife86]

Works good! Love the new update!

One thing might think of is the binding port 8080, I for some reason already have that one bound to something else so I have to change the docker-compose so i can use 8088. Maybe make that a variable. Because of this I cant use the pull images feature.

If not then its not a worry. I can manage my local docker-compose file.

Would it be possible to add a custom service via docker hub name and tag? Just a thought.

[GeiserX]

Keep it coming with such awesome ideas @LiquidLife86 I think these are good kind of "needs". I expect other users to face this too. Incoming patch.

[GeiserX]

Both are in.

Port. The compose file now reads the host port from CASHPILOT_UI_PORT. Put CASHPILOT_UI_PORT=8088 in the .env next to docker-compose.yml, then docker compose up -d. The in-app update keeps working on 8088: before it replaces anything it reads your compose file, and it refuses if the new container would publish different ports. Your own edited 8088:8080 line is accepted too. It ships in the next release; UPGRADING.md has the details.
#561

Custom services. Catalog, Your custom services, Add custom service. You give a name, an image with its tag (Docker Hub or any registry), environment variables (tick Secret for tokens), an optional command and data path, and a memory limit. It deploys to any of your boxes and shows on the dashboard as Custom, with no balance, because there is no earnings collector for it. It runs with the same lockdown as everything else: all capabilities dropped, no privileged mode, no host network, no host folders and no published ports. So an image that needs any of those, like a systemd one, won't start, and you get a clear error instead of a restart loop. No tag is accepted with a warning, because latest changes under you.
#562

Thanks again, these two came straight from your setup.