Modify presubmits to support testing with v1alpha2 (kubeflow#632)

* Changes to support v1alpha2 testing in presubmits. * The tests are currently disabled because they aren't passing yet because termination policy isn't handled correctly (kubeflow#634) * Changed the v1alpha2 test to use the same smoke test as used by v1alpha1 as opposed to using mnist. mnist causing problems because of issues downloading the data see kubeflow/kubeflow#974 * We want a simpler test that allows for more direct testing of the distributed communication pattern * Also mnist is expensive in that it tries to download data. * Add a parameter tfJobVersion to the deploy script so we can control whether we deploy v1alpha1 or v1alpha2 * Parameterize the E2E test workflow by the TFJob version we want to run. * update test-app - We need to pull in a version of the app which has the TFJobVersion flag. * Create a script to regenerate the test-app for future use. Related to kubeflow#589 * Fix versionTag logic; we need to allow for case where versionTag is an empty string.
yph152 · Jun 18, 2018 · af9f990 · af9f990
1 parent 3900ad8
commit af9f990
Show file tree

Hide file tree

Showing 61 changed files with 3,921 additions and 151,815 deletions.
diff --git a/prow_config.yaml b/prow_config.yaml
@@ -4,3 +4,13 @@ workflows:
   - app_dir: kubeflow/tf-operator/test/workflows
     component: workflows
     name: tfjob-e2e
+    params:
+      tfJobVersion: v1alpha1
+  # TODO*https://github.com/kubeflow/tf-operator/issues/634)
+  # Enable the v1alpha2 once the job successfully completes
+  # when master completes.
+  #- app_dir: kubeflow/tf-operator/test/workflows
+  #  component: workflows
+  #  name: tfjob-e2e-v1alpha2
+  #  params:
+  #    tfJobVersion: v1alpha2
diff --git a/py/deploy.py b/py/deploy.py
@@ -152,6 +152,7 @@ def setup(args):
       "tfJobImage": args.image,
       "name": "kubeflow-core",
       "namespace": args.namespace,
+      "tfJobVersion":  args.tf_job_version,
     }
 
     component = "core"
@@ -171,8 +172,15 @@ def setup(args):
     util.setup_cluster(api_client)
 
     # Verify that the TfJob operator is actually deployed.
-    tf_job_deployment_name = "tf-job-operator"
-    logging.info("Verifying TfJob controller started.")
+    if args.tf_job_version == "v1alpha1":
+      tf_job_deployment_name = "tf-job-operator"
+    elif args.tf_job_version == "v1alpha2":
+      tf_job_deployment_name = "tf-job-operator-v1alpha2"
+    else:
+      raise ValueError(
+        "Unrecognized value for tf_job_version %s" % args.tf_job_version)
+    logging.info("Verifying TfJob deployment %s started.",
+                 tf_job_deployment_name)
 
     # TODO(jlewi): We should verify the image of the operator is the correct.
     util.wait_for_deployment(api_client, args.namespace, tf_job_deployment_name)
@@ -247,6 +255,11 @@ def main():  # pylint: disable=too-many-locals
     action="append",
     help="Accelerator to add to the cluster. Should be of the form type=count.")
 
+  parser_setup.add_argument(
+    "--tf_job_version",
+    dest="tf_job_version",
+    help="Which version of the TFJobOperator to deploy.")
+
   parser_setup.set_defaults(func=setup)
   add_common_args(parser_setup)
 

diff --git a/test/recreate_app.sh b/test/recreate_app.sh
@@ -0,0 +1,50 @@
+#!/bin/bash
+#
+# A simple script to recreate the Kubeflow test app
+#
+set -ex
+# Create a namespace for kubeflow deployment
+NAMESPACE=kubeflow
+
+# Which version of Kubeflow to use
+# For a list of releases refer to:
+# https://github.com/kubeflow/kubeflow/releases
+VERSION=master
+API_VERSION=v1.7.0
+
+DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
+
+cd ${DIR}
+
+APP_NAME=test-app
+
+
+if [ -d ${DIR}/${APP_NAME} ]; then
+	# TODO(jlewi): Maybe we should prompt to ask if we want to delete?
+	echo "Directory ${DIR}/${APP_NAME} exists"
+	echo "Do you want to delete ${DIR}/${APP_NAME} y/n[n]:"
+	read response
+
+	if [ "${response}"=="y" ]; then
+		rm -r ${DIR}/${APP_NAME}
+	else
+		"Aborting"
+		exit 1
+	fi
+fi
+
+ks init ${APP_NAME} --api-spec=version:${API_VERSION}
+cd ${APP_NAME}
+ks env set default --namespace ${NAMESPACE}
+
+# Install Kubeflow components
+ks registry add kubeflow github.com/kubeflow/kubeflow/tree/${VERSION}/kubeflow
+
+ks pkg install kubeflow/core@${VERSION}
+
+# Create templates for core components
+ks generate kubeflow-core core
+
+# Run autoformat from the git root
+cd ${DIR}/..
+bash <(curl -s https://raw.githubusercontent.com/kubeflow/kubeflow/${VERSION}/scripts/autoformat_jsonnet.sh)
diff --git a/test/test-app/.gitignore b/test/test-app/.gitignore
@@ -0,0 +1,4 @@
+/lib
+/.ksonnet/registries
+/app.override.yaml
+/.ks_environment
diff --git a/test/test-app/.ksonnet/registries/incubator/422d521c05aa905df949868143b26445f5e4eda5.yaml b/test/test-app/.ksonnet/registries/incubator/422d521c05aa905df949868143b26445f5e4eda5.yaml
diff --git a/test/test-app/.ksonnet/registries/kubeflow/845f2a02e6ef4e25cae8555a37924d3510d07b36.yaml b/test/test-app/.ksonnet/registries/kubeflow/845f2a02e6ef4e25cae8555a37924d3510d07b36.yaml
diff --git a/test/test-app/app.yaml b/test/test-app/app.yaml
@@ -2,29 +2,29 @@ apiVersion: 0.1.0
 environments:
   default:
     destination:
-      namespace: kubeflow-releasing
-      server: https://35.226.49.107
+      namespace: kubeflow
+      server: https://35.196.213.148
     k8sVersion: v1.7.0
     path: default
 kind: ksonnet.io/app
 libraries:
   core:
     gitVersion:
-      commitSha: 845f2a02e6ef4e25cae8555a37924d3510d07b36
+      commitSha: f7a68336ad7a65c2cbba8462e89d24a10626687e
       refSpec: master
     name: core
     registry: kubeflow
 name: test-app
 registries:
   incubator:
     gitVersion:
-      commitSha: 422d521c05aa905df949868143b26445f5e4eda5
+      commitSha: 40285d8a14f1ac5787e405e1023cf0c07f6aa28c
       refSpec: master
     protocol: github
     uri: github.com/ksonnet/parts/tree/master/incubator
   kubeflow:
     gitVersion:
-      commitSha: 845f2a02e6ef4e25cae8555a37924d3510d07b36
+      commitSha: f7a68336ad7a65c2cbba8462e89d24a10626687e
       refSpec: master
     protocol: github
     uri: github.com/kubeflow/kubeflow/tree/master/kubeflow

diff --git a/test/test-app/components/core.jsonnet b/test/test-app/components/core.jsonnet
@@ -1,5 +1,13 @@
+local env = std.extVar("__ksonnet/environments");
 local params = std.extVar("__ksonnet/params").components.core;
-local k = import 'k.libsonnet';
+
+local k = import "k.libsonnet";
 local all = import "kubeflow/core/all.libsonnet";
 
-std.prune(k.core.v1.list.new(all.parts(params).all))
+// updatedParams uses the environment namespace if
+// the namespace parameter is not explicitly set
+local updatedParams = params {
+  namespace: if params.namespace == "null" then env.namespace else params.namespace,
+};
+
+std.prune(k.core.v1.list.new(all.parts(updatedParams).all))
diff --git a/test/test-app/components/params.libsonnet b/test/test-app/components/params.libsonnet
@@ -10,13 +10,22 @@
       cloud: "null",
       disks: "null",
       jupyterHubAuthenticator: "null",
-      jupyterHubImage: "gcr.io/kubeflow/jupyterhub-k8s:1.0.1",
+      jupyterHubImage: "gcr.io/kubeflow/jupyterhub-k8s:v20180531-3bb991b1",
       jupyterHubServiceType: "ClusterIP",
+      jupyterNotebookPVCMount: "null",
+      jupyterNotebookRegistry: "gcr.io",
+      jupyterNotebookRepoName: "kubeflow-images-public",
       name: "core",
-      namespace: "default",
+      namespace: "null",
+      reportUsage: "false",
+      tfAmbassadorImage: "quay.io/datawire/ambassador:0.30.1",
+      tfAmbassadorServiceType: "ClusterIP",
       tfDefaultImage: "null",
-      tfJobImage: "gcr.io/tf-on-k8s-dogfood/tf_operator:v20180131-cabc1c0-dirty-e3b0c44",
+      tfJobImage: "gcr.io/kubeflow-images-public/tf_operator:v20180522-77375baf",
       tfJobUiServiceType: "ClusterIP",
+      tfJobVersion: "v1alpha1",
+      tfStatsdImage: "quay.io/datawire/statsd:0.30.1",
+      usageId: "unknown_cluster",
     },
   },
 }
diff --git a/test/test-app/environments/default/.metadata/k.libsonnet b/test/test-app/environments/default/.metadata/k.libsonnet