Update tekton task to Platform API 0.2

[jromero]

Update tekton template to adhere to Platform API version 0.2.

Additional Considerations

• We should use the sample builder and sample java app as inputs on the example to make it easier to use off-the-shelf while annotating what and how things should be configured. This is in relation to a conversation on slack and https://github.com/tektoncd/catalog/issues/151.

Acceptance Criteria

AC1. Platform API verification

Given the tekton task is configured for Platform API version 0.2
When I provide a builder with lifecycle implementing to Platform API version 0.3
Then task should fail with error message:
`Lifecycle's Platform API version is 0.3 which is incompatible with Platform API version 0.2`

Notes

• Template is housed: https://github.com/tektoncd/catalog/tree/master/buildpacks

Relates to:

• https://github.com/buildpacks/docs/issues/111

Blocked by:

• https://github.com/buildpack/rfcs/pull/29

[zmackie]

Is there any sort of versioning associated with the template? Maybe that's something we should do (if possible).

[jromero]

I was thinking about this... We could create a yaml file per platform API version (or keep older versions in a directory). Otherwise I can't think of a versioning mechanism since it's not our repository.

[natalieparellano]

Acceptance steps:

• Setup a local k8s cluster
  • Docker > Preferences > Kubernetes > Enable Kubernetes > Apply & Restart
  • kubectl config current-context should be docker-desktop
• Run a local registry
  • docker run --detach -p5000:5000 registry:2
• Permit insecure registry:
  • Docker > Preferences > Docker Engine > add the line "insecure-registries" : ["localhost:5000"] > Apply & Restart
• Install tekton
  • kubectl apply --filename https://storage.googleapis.com/tekton-releases/pipeline/latest/release.yaml
• Install buildpacks task
  • kubectl apply -f ./buildpacks-v3.yaml
• Apply taskrun
  • Edit the example under Usage in the README
  • output url will look like <your external ip>:5000/some-output-image
  • kubectl apply -f <path to task run yaml>
• See task succeed
• See output image written to registry
  • docker pull localhost:5000/some-output-image

To test persistent cache:

• Create a persistent volume pointing to local cache directory
  • mkdir <some dir>
  • ensure the directory is writable by the world: chmod 777 <some dir>
  • edit pv-volume.yaml to have the correct host path
  • kubectl apply -f <path to volume yaml>
• Create a persistent volume claim
  • kubectl apply -f https://k8s.io/examples/pods/storage/pv-claim.yaml
• Update the taskrun
  • Uncomment the lines that pertain to caching
• Apply taskrun
  • kubectl apply -f <path to task run yaml>
• See task succeed
• See items in the cache directory

cc @micahyoung

[dfreilich]

@natalieparellano Was this taken care of?

[natalieparellano]

Yep!