(merge dev branch) refactor, chore: refactor charm to use `Deployment…

…` for workload, also bumps training-operator 1.7->1.8 (#167)

* pin integration test dependencies, refactor constants in tests (#164)
* refactor: deploy the training-operator with kubernetes resources (#161)
* chore: bump training-operator v1.7 -> v1.8 (#162)
* refactor: apply a workload Service instead of using juju created one (#173)
* tests: skip test_upgrade due to #170 (#171)
* build, tests: bump charmed-kubeflow-chisme 0.4.0 -> 0.4.1 (#172)

Fixes #159

.github/workflows/integrate.yaml

@@ -75,8 +75,20 @@ jobs:
         run: juju status
         if: failure()
 
-      - name: Get workload logs
-        run: kubectl logs --tail 100 -ntesting -lapp.kubernetes.io/name=training-operator -ccharm
+      - name: Get validatingwebhookconfigurations
+        run: kubectl get validatingwebhookconfigurations validator.training-operator.kubeflow.org -oyaml
+        if: failure()
+
+      - name: Get secret
+        run: kubectl get secret -ntesting training-operator-webhook-cert -oyaml
+        if: failure()
+
+      - name: Describe pod
+        run: kubectl describe pod -ntesting -lapp.kubernetes.io/name=training-operator
+        if: failure()
+
+      - name: Get pods
+        run: kubectl get pods -A
         if: failure()
 
       - name: Get operator logs

charmcraft.yaml

@@ -12,4 +12,3 @@ bases:
 parts:
   charm:
     charm-python-packages: [setuptools, pip]
-    build-packages: [git]

config.yaml

@@ -0,0 +1,6 @@
+options:
+  training-operator-image:
+    default: kubeflow/training-operator:v1-4485b0a
+    description: |
+      Container image to be used by the training-operator workload.
+    type: string

examples/paddlejob.yaml

@@ -1,3 +1,4 @@
+# From https://github.com/kubeflow/training-operator/blob/v1.8-branch/examples/paddlepaddle/simple-cpu.yaml
 apiVersion: "kubeflow.org/v1"
 kind: PaddleJob
 metadata:
@@ -21,4 +22,4 @@ spec:
               ports:
                 - containerPort: 37777
                   name: master
-              imagePullPolicy: Always
+              imagePullPolicy: Always

examples/pytorchjob.yaml

@@ -1,3 +1,4 @@
+# From https://github.com/kubeflow/training-operator/blob/v1.8-branch/examples/pytorch/simple.yaml
 apiVersion: "kubeflow.org/v1"
 kind: PyTorchJob
 metadata:

examples/tfjob.yaml

@@ -1,17 +1,18 @@
+# From https://github.com/kubeflow/training-operator/blob/v1.8-branch/examples/tensorflow/simple.yaml
 apiVersion: "kubeflow.org/v1"
 kind: TFJob
 metadata:
   name: tfjob-simple
 spec:
-   tfReplicaSpecs:
-     Worker:
-       replicas: 2
-       restartPolicy: OnFailure
-       template:
-         spec:
-           containers:
-             - name: tensorflow
-               image: gcr.io/kubeflow-ci/tf-mnist-with-summaries:1.0
-               command:
-                 - "python"
-                 - "/var/tf_mnist/mnist_with_summaries.py"
+  tfReplicaSpecs:
+    Worker:
+      replicas: 2
+      restartPolicy: OnFailure
+      template:
+        spec:
+          containers:
+            - name: tensorflow
+              image: kubeflow/tf-mnist-with-summaries:latest
+              command:
+                - "python"
+                - "/var/tf_mnist/mnist_with_summaries.py"

examples/xgboostjob.yaml

@@ -1,7 +1,8 @@
+# From https://github.com/kubeflow/training-operator/blob/v1.8-branch/examples/xgboost/xgboostjob.yaml
 apiVersion: kubeflow.org/v1
 kind: XGBoostJob
 metadata:
-  name: xgboost-simple
+  name: xgboost-dist-iris-test-train
 spec:
   xgbReplicaSpecs:
     Master:
@@ -11,7 +12,7 @@ spec:
         spec:
           containers:
           - name: xgboost
-            image: docker.io/merlintang/xgboost-dist-iris:1.1
+            image: docker.io/kubeflow/xgboost-dist-iris:latest
             ports:
             - containerPort: 9991
               name: xgboostjob-port
@@ -30,7 +31,7 @@ spec:
         spec:
           containers:
           - name: xgboost
-            image: docker.io/merlintang/xgboost-dist-iris:1.1
+            image: docker.io/kubeflow/xgboost-dist-iris:latest
             ports:
             - containerPort: 9991
               name: xgboostjob-port