You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It could potentially be Dokploy's fault, I'm not really sure if the worker should have it's separate container, or it's running in the same container with the web process. If it does: provide how to verify it.
I will provide any other information you will ask.
Already tried:
Saved the Procfile in ASCI encoding (to remove potential BOM before my processes).
Used sh -c 'worker command'.
Removed the web process to verify worker process - it runs, creates a container in which the worker lives, like it should be.
The text was updated successfully, but these errors were encountered:
Looking through the deploy log it seems like everything is in order. I'm not familiar with Dokploy so I can't say for sure if something is off with how they handle the entrypoints created by the Procfile buildpack in the container image that's produced.
Here are some steps you can use to locally verify what's produced by a Cloud Native Buildpack:
Inspect the container image produced with pack inspect:
pack inspect <your-image-name>
This will show you metadata such as the builder + buildpacks that produced the container, the run image that will be used, and the registered process types.
Start the container using the default (web) process type to confirm it works as expected:
docker run --rm <your-image-name>
Start the container using the worker process type to confirm it works as expected:
docker run --rm --entrypoint worker <your-image-name>
If the steps above work for you, then the buildpack and container image are working as expected. You'll notice that only one container image is produced which contains both the web and worker processes. If Dokploy does require separate containers for each process then maybe file an issue with them to see how to best support this use case and link it here. I'd be interested to hear how CNBs could run better on their platform.
Also, is there a reason why you needed to add export PATH="/layers/heroku_nodejs-engine/dist/bin:$PATH" to your web process? The CNB-provided launcher that executes processes types will automatically set environment variables configured by buildpacks and the node-engine buildpack already adds its binaries to the PATH.
I have a project on NextJS (14.2.21), running on Dokploy (0.17.9), deploying on Heroku Buildpacks.
Inside of my project I have the next Procfile:
At build, the web process is starting well, but the worker is not appearing (should be a separate container excplicitelly for the worker).
Providing my deployment log: deploy log
It could potentially be Dokploy's fault, I'm not really sure if the worker should have it's separate container, or it's running in the same container with the web process. If it does: provide how to verify it.
I will provide any other information you will ask.
Already tried:
The text was updated successfully, but these errors were encountered: