Surface spin up errors?

[kate-goldenring]

When I run a spin app that requires two application variables but only provide one, the app fails to start up but nothing about the Pod events explains why:

  Normal   Created    40s (x4 over 85s)  kubelet            Created container spin-foo-bar-variables
  Normal   Started    40s (x4 over 84s)  kubelet            Started container spin-foo-bar-variables
  Warning  BackOff    3s (x8 over 83s)   kubelet            Back-off restarting failed container spin-foo-bar-variables in pod spin-foo-bar-variables-7858fcfb45-7lmgq_default(4cabc0f7-5d02-4906-9296-fc250dd6ffe9)

Logs for the container cannot be retrieved. I would expect to see an error there. Locally, a spin up without a required error would throw this error:

Error: provider error: no provider resolved required variable "foo"

Caused by:
      no provider resolved required variable "foo"

Reproduce

Apply this app using the Spin operator which expects foo and bar variables to be defined

apiVersion: core.spinoperator.dev/v1alpha1
kind: SpinApp
metadata:
  name: spin-foo-bar-variables
spec:
  image: "ghcr.io/kate-goldenring/spin-foo-bar-variables:v1"
  executor: containerd-shim-spin
  replicas: 1
  variables:
  - name: bar
    value: yoo

[Mossaka]

I think this is because all the spin executor errors are treated the same as shim errors, so they are being piped to containerd log. This is not desirable, we want to be able to pipe shim executor errors to Pod errors.