fix(test): Don't tear down AirflowClusters when --skip-delete is used

tests/templates/kuttl/ca-cert/90-cleanup-executor-pods.yaml

@@ -1,18 +1,8 @@
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster --all -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh

tests/templates/kuttl/commons/cleanup-executor-pods.sh

@@ -0,0 +1,32 @@
+#!/usr/bin/env sh
+# Shared teardown for KubernetesExecutor tests: force-deletes leftover DAG task pods so
+# that kuttl can delete the namespace within its timeout. Their Vector sidecar ignores
+# SIGTERM (it is not PID 1) and would linger for the full 300s grace period.
+# The proper fix is in operator-rs (making Vector PID 1 via exec).
+
+# With --skip-delete the namespace is kept for debugging, so this must not run. kuttl does
+# not expose the flag to test steps, hence looking for it in the argv of the kuttl ancestor.
+pid=$(ps -o ppid= -p "$$" 2>/dev/null | tr -d '[:space:]')
+while [ "${pid:-0}" -ge 1 ]; do
+  args=$(ps -o args= -p "$pid" 2>/dev/null)
+  case "$args" in
+    *kuttl*)
+      case "$args" in
+        *--skip-delete*)
+          echo "kuttl was started with --skip-delete, keeping the AirflowCluster"
+          exit 0
+          ;;
+      esac
+      ;;
+  esac
+  pid=$(ps -o ppid= -p "$pid" 2>/dev/null | tr -d '[:space:]')
+done
+
+# shellcheck disable=SC2154 # NAMESPACE is set by kuttl for test-step commands
+kubectl delete airflowcluster --all -n "$NAMESPACE" --wait=false 2>/dev/null || true
+
+if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n "$NAMESPACE" --timeout=120s 2>/dev/null; then
+  exit 0
+fi
+kubectl delete pods -l app.kubernetes.io/name=airflow -n "$NAMESPACE" --grace-period=0 --force 2>/dev/null || true
+kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n "$NAMESPACE" --timeout=300s

tests/templates/kuttl/external-access/90-cleanup-executor-pods.yaml.j2

@@ -1,20 +1,10 @@
 {% if test_scenario['values']['executor'] == 'kubernetes' %}
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh
 {% endif %}

tests/templates/kuttl/ldap/99-cleanup-executor-pods.yaml.j2

@@ -1,20 +1,10 @@
 {% if test_scenario['values']['executor'] == 'kubernetes' %}
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh
 {% endif %}

tests/templates/kuttl/logging/90-cleanup-executor-pods.yaml.j2

@@ -1,20 +1,10 @@
 {% if test_scenario['values']['executor'] == 'kubernetes' %}
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh
 {% endif %}

tests/templates/kuttl/mount-dags-configmap/90-cleanup-executor-pods.yaml.j2

@@ -1,20 +1,10 @@
 {% if test_scenario['values']['executor'] == 'kubernetes' %}
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh
 {% endif %}

tests/templates/kuttl/mount-dags-gitsync/90-cleanup-executor-pods.yaml.j2

@@ -1,20 +1,10 @@
 {% if test_scenario['values']['executor'] == 'kubernetes' %}
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh
 {% endif %}

tests/templates/kuttl/oidc/90-cleanup-executor-pods.yaml

@@ -1,18 +1,8 @@
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh

tests/templates/kuttl/opa/90-cleanup-executor-pods.yaml

@@ -1,18 +1,8 @@
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh

tests/templates/kuttl/overrides/90-cleanup-executor-pods.yaml

@@ -1,18 +1,8 @@
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster --all -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh

tests/templates/kuttl/remote-logging/90-cleanup-executor-pods.yaml.j2

@@ -1,20 +1,10 @@
 {% if test_scenario['values']['executor'] == 'kubernetes' %}
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh
 {% endif %}

tests/templates/kuttl/smoke/90-cleanup-executor-pods.yaml.j2

@@ -1,25 +1,10 @@
 {% if test_scenario['values']['executor'] == 'kubernetes' %}
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
-#
-# FIXME: This step also runs when kuttl is called with --skip-delete, so the
-# AirflowCluster is torn down even though the namespace is kept for debugging.
-# kuttl does not expose --skip-delete to test steps, so there is no clean way to
-# guard against it from here.
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh
 {% endif %}

tests/templates/kuttl/triggerer/90-cleanup-executor-pods.yaml.j2

@@ -1,20 +1,10 @@
 {% if test_scenario['values']['executor'] == 'kubernetes' %}
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh
 {% endif %}

tests/templates/kuttl/versioning/90-cleanup-executor-pods.yaml

@@ -1,18 +1,8 @@
 ---
-# Force-delete KubernetesExecutor DAG task pods before kuttl deletes the namespace.
-# Their Vector sidecar does not respond to SIGTERM (it is not PID 1), so these pods
-# sit in Terminating for the full terminationGracePeriodSeconds (300s), blocking
-# namespace deletion past kuttl's timeout.
-# The proper fix is in operator-rs (making Vector PID 1 via exec).
+# Force-delete leftover KubernetesExecutor DAG task pods so that kuttl can delete the
+# namespace within its timeout. See the script for the details.
 apiVersion: kuttl.dev/v1beta1
 kind: TestStep
 timeout: 600
 commands:
-  - script: |
-      kubectl delete airflowcluster airflow -n $NAMESPACE --wait=false 2>/dev/null || true
-  - script: |
-      if kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=120s 2>/dev/null; then
-        exit 0
-      fi
-      kubectl delete pods -l app.kubernetes.io/name=airflow -n $NAMESPACE --grace-period=0 --force 2>/dev/null || true
-      kubectl wait --for=delete pod -l app.kubernetes.io/name=airflow -n $NAMESPACE --timeout=300s
+  - script: sh ../../../../templates/kuttl/commons/cleanup-executor-pods.sh