What happened?

brian@brian-LNQ0KYW6HD:devpod-provider-nomad λ docker ps -a                                                                                                                          ±[main]
CONTAINER ID   IMAGE     COMMAND            CREATED         STATUS         PORTS     NAMES
15b1ef470ea1   alpine    "sleep infinity"   3 minutes ago   Up 3 minutes             devpod-258d6de8-12e9-3064-a81f-c92fb7dc9cd6
brian@brian-LNQ0KYW6HD:devpod-provider-nomad λ cat .devcontainer.json                                                                                                                ±[main]
{"image":"mcr.microsoft.com/devcontainers/go","build":{}}

The create command doesn't seem to be able to load the driver configuration from Devpod, so it launches with the default command and image rather than reading from devcontainer.json

What did you expect to happen instead?

Use the image defined in the devcontainer.

How can we reproduce the bug? (as minimally and precisely as possible)

My devcontainer.json:

{"image":"mcr.microsoft.com/devcontainers/go","build":{}}

Local Environment:

• DevPod Version: [use devpod version]
• Operating System: windows | linux | mac
• ARCH of the OS: AMD64 | ARM64 | i386
• Nomad version [use nomad version]

DevPod Provider:

• Custom provider: provide imported provider.yaml config file

Anything else we need to know?

Add more devpod configuration options to use for configuring the Nomad job we launch.

What happened?

brian@brian-LNQ0KYW6HD:devpod-provider-nomad λ devpod ssh --debug                                                                                                                    ±[main]

? Please select a workspace from the list below devpod-provider-nomad
11:33:19 debug Adding ssh keys to agent, disable via 'devpod context set-options -o SSH_ADD_PRIVATE_KEYS=false'
11:33:19 debug Acquire workspace lock...
11:33:19 debug Acquired workspace lock...
11:33:19 debug Acquire machine lock...
11:33:19 debug Acquired machine lock...
11:33:19 debug Run status provider command: ${NOMAD_PROIDER} status
11:33:19 fatal get status: bash: status: command not found

I think this is because only machine based providers are supposed to have a status command. Since this is not a machine provider, maybe remove the status command from the provider config and it'll work?

What did you expect to happen instead?

Can properly ssh.

How can we reproduce the bug? (as minimally and precisely as possible)

n/a, any devcontainer should work.

Local Environment:

• DevPod Version: [use devpod version]
• Operating System: windows | linux | mac
• ARCH of the OS: AMD64 | ARM64 | i386
• Nomad version [use nomad version]

DevPod Provider:

• Custom provider: provide imported provider.yaml config file

Anything else we need to know?

Properly implement the devpod command command. Maybe use ActionExec instead of the raw nomad CLI nomad alloc exec?

https://devpod.sh/docs/managing-providers/add-provider#community-providers

Once this is more functional, add it to the list.