Based on SCDF Server Kubernetes

I.e. add

...
- GET    /resources/maven/**             => hasRole('ROLE_VIEW')

to dataflow-server.yml

Hi, I am following this tutorial about Spring Cloud Data Flow with Nomad:

https://donovanmuller.github.io/spring-cloud-dataflow-server-nomad/docs/1.1.0.RELEASE/reference/htmlsingle/#_download_and_run_the_spring_cloud_data_flow_shell

And I've started after "9 Getting Started", the part with Vagrant goes fine and my VM is functioning. The problem is executing:

nomad run https://raw.githubusercontent.com/donovanmuller/spring-cloud-dataflow-server-nomad/v1.1.0.RELEASE/src/etc/nomad/nexus.nomad

That I've changed to:

nomad run https://raw.githubusercontent.com/donovanmuller/spring-cloud-dataflow-server-nomad/v1.1.0.RELEASE/src/etc/nomad/scdf.nomad

All services goes fine, excluding scdf-server, which image can be found here. Consul always get 503 { Status: Down} from http://scdf-server.hashistack.vagrant, but this is not the crucial problem. Ignoring the health check, after that I've executed:

Which gives me "server-unknown". After a bit digging in the scdf.nomad file, I've reached the http://172.16.0.2:9393, that is the IP running my scdf-server, but the address http://scdf-server.hashistack.vagrant does not respond.

After discovering the IP, I've executed:

java -jar spring-cloud-dataflow-shell-1.1.0.RELEASE.jar --dataflow.uri=http://172.16.0.2:9393, which was a success. But, when continuing with the tutorial, when I try to execute:

dataflow:> runtime apps

I got connection refused, and can't proceed anymore. I don't have any knowledge about the docker image and I'm really lost here. I have a few questions about this. Am i doing something wrong? I've already changed the .nomad file a thousand times and tried a lot but I'm still stuck.

Nomad 0.6.x changed the handling of stopped Jobs, which results in the deployer status being invalid in some cases.

Previously stopped Job statuses were removed. Now, stopped Jobs have a new stopped state.
See hashicorp/nomad#2566

Need to update the Nomad jobs on every release.

Probably should be automated as part of the mvn release to keep in sync.

See https://www.nomadproject.io/docs/job-specification/template.html

This could enable support for storing a database password as a Vault secret and have the template use it.

https://github.com/donovanmuller/spring-cloud-deployer-nomad/releases/tag/v1.2.3.RELEASE

See structure of Spring Cloud Data Flow - Kubernetes: https://github.com/spring-cloud/spring-cloud-dataflow-server-kubernetes

Do you plan to support the "Raw Fork/Exec Driver".
It is unclear why someone thinks that raw_exec driver (without isolation) is a bad thing.
The old SpringXD execution mode should have been preserved (i.e just run microservice(jvm, python, whatever)) like any other program natively). The "microservices" idea is OK but let us decide whether we we want to use docker or not.
Supervised Docker environment (like Kubernetes which I have tested) is unnecessary overhead for some use-cases. For example: I'm not a big scale service provider", I just want to run my small distributed data pipeline.

Thanks in advance

Best regards,
Ivan Prostran