nodeSelector value not propagated to child pipelines in pipeline-in-pipeline feature

[gigatesseract]

Expected Behavior

When running a nested pipeline configuration, the node selector value is not passed to child pipelines. This results in nodeSelector being None.

Actual Behavior

Node selector to be passed to child pipelines.

Steps to Reproduce the Problem

1. Ensure pipeline-in-pipeline is enabled. (feature flag set to True and pod running in tekton-pipelines namespace)
2. Apply the following YAMLs in the same namespace
3. Create a custom node selector

apiVersion: tekton.dev/v1beta1
kind: Task
metadata:
  name: hello-world
spec:
  steps:
    - name: echo
      image: alpine
      script: |
        #!/bin/sh
        echo "Hello World"
---
apiVersion: tekton.dev/v1beta1
kind: Pipeline
metadata:
  name: pipeline-in-pipeline-test
spec:
  tasks:
    - name: custom-task-inside-pipeline-in-pipeline
      taskRef:
        kind: Task
        name: hello-world
      timeout: 48h10m0s
---
apiVersion: tekton.dev/v1beta1
kind: Pipeline
metadata:
  name: test-node-selector
spec:
  tasks:
    - name: spawned-directly
      taskRef:
        kind: Task
        name: hello-world
    - name: indirect-pipeline-in-pipeline-test
      taskRef:
        apiVersion: tekton.dev/v1beta1
        kind: Pipeline
        name: pipeline-in-pipeline-test
      timeout: 24h0m0s

4. Create the run yaml

apiVersion: tekton.dev/v1beta1
kind: PipelineRun
metadata:
  generateName: test-node-selector-run-
spec:
  pipelineRef:
    name: test-node-selector
  podTemplate:
    nodeSelector:
      kubernetes.io/role: <CUSTOM_NODE_SELECTOR>
  timeout: 24h0m0s

5. Describe the pod created for the task which was created for the nested pipeline. We get something like

QoS Class:                   BestEffort
Node-Selectors:              <none>
Tolerations:                 node.kubernetes.io/not-ready:NoExecute op=Exists for 300s

Additional Info

• Kubernetes version:

  Output of kubectl version:

Client Version: version.Info{Major:"1", Minor:"23", GitVersion:"v1.23.3", GitCommit:"816c97ab8cff8a1c72eccca1026f7820e93e0d25", GitTreeState:"clean", BuildDate:"2022-01-25T21:25:17Z", GoVersion:"go1.17.6", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"25", GitVersion:"v1.25.11", GitCommit:"8cfcba0b15c343a8dc48567a74c29ec4844e0b9e", GitTreeState:"clean", BuildDate:"2023-06-14T09:49:38Z", GoVersion:"go1.19.10", Compiler:"gc", Platform:"linux/amd64"}
WARNING: version difference between client (1.23) and server (1.25) exceeds the supported minor version skew of +/-1```

- Tekton Pipeline version:


  **Output of `tkn version` or `kubectl get pods -n tekton-pipelines -l app=tekton-pipelines-controller -o=jsonpath='{.items[0].metadata.labels.version}'`**

(paste your output here)

I dont have access to list pods in `tekton-pipelines` namespace, we have the dashboard UI and the version from that is
`v0.47.2`


<!-- Any other additional information -->
Can confirm that this issue is not present in `v0.35.1` (We have clusters running tekton pipelines of both these versions and issue is only in 0.47.2)

[gigatesseract]

The issue seems to be here: https://github.com/tektoncd/experimental/tree/main/pipelines-in-pipelines/pkg/reconciler/pip/piprun.go in function ReconcileKind, where check if the CustomRun has started, and if not, start the run.

However, I expect the spec to be copied as is from parent pipeline run to the custom run. I am new to go code, can you perhaps point me to the place in pipelines-in-pipelines where we copy the spec of the pipelinerun run as is to the custom run?

[sallyyama]

I opened the PR to address this issue. https://github.com/tektoncd/experimental/pull/966