Fix kind mode and add k8s deployment test (#704)

* Fix kind mode and add k8s deployment test

* Fix lint errors

stack_orchestrator/deploy/k8s/cluster_info.py

@@ -168,8 +168,8 @@ class ClusterInfo:
             result.append(pv)
         return result
 
-    # to suit the deployment, and also annotate the container specs to point at said volumes
+    # TODO: put things like image pull policy into an object-scope struct
-    def get_deployment(self):
+    def get_deployment(self, image_pull_policy: str = None):
         containers = []
         for pod_name in self.parsed_pod_yaml_map:
             pod = self.parsed_pod_yaml_map[pod_name]
@@ -189,7 +189,7 @@ class ClusterInfo:
                 container = client.V1Container(
                     name=container_name,
                     image=image_to_use,
-                    image_pull_policy="Always",
+                    image_pull_policy=image_pull_policy,
                     env=envs_from_environment_variables_map(self.environment_variables.map),
                     ports=[client.V1ContainerPort(container_port=port)],
                     volume_mounts=volume_mounts,

stack_orchestrator/deploy/k8s/deploy_k8s.py

@@ -111,7 +111,7 @@ class K8sDeployer(Deployer):
                 print("PVCs created:")
                 print(f"{pvc_resp}")
         # Process compose files into a Deployment
-        deployment = self.cluster_info.get_deployment()
+        deployment = self.cluster_info.get_deployment(image_pull_policy=None if self.is_kind() else "Always")
         # Create the k8s objects
         if opts.o.debug:
             print(f"Sending this deployment: {deployment}")
@@ -132,7 +132,7 @@ class K8sDeployer(Deployer):
             print("Service created:")
             print(f"{service_resp}")
 
-        # TODO: disable ingress for kind
+        if not self.is_kind():
             ingress: client.V1Ingress = self.cluster_info.get_ingress()
 
             if opts.o.debug:
@@ -196,7 +196,7 @@ class K8sDeployer(Deployer):
         except client.exceptions.ApiException as e:
             _check_delete_exception(e)
 
-        # TODO: disable ingress for kind
+        if not self.is_kind():
             ingress: client.V1Ingress = self.cluster_info.get_ingress()
             if opts.o.debug:
                 print(f"Deleting this ingress: {ingress}")
@@ -219,7 +219,7 @@ class K8sDeployer(Deployer):
 
         if all_pods.items:
             for p in all_pods.items:
-                if self.cluster_info.app_name in p.metadata.name:
+                if f"{self.cluster_info.app_name}-deployment" in p.metadata.name:
                     pods.append(p)
 
         if not pods:
@@ -266,7 +266,7 @@ class K8sDeployer(Deployer):
         ret = []
 
         for p in pods.items:
-            if self.cluster_info.app_name in p.metadata.name:
+            if f"{self.cluster_info.app_name}-deployment" in p.metadata.name:
                 pod_ip = p.status.pod_ip
                 ports = AttrDict()
                 for c in p.spec.containers:
@@ -299,11 +299,20 @@ class K8sDeployer(Deployer):
 
     def logs(self, services, tail, follow, stream):
         self.connect_api()
-        pods = pods_in_deployment(self.core_api, "test-deployment")
+        pods = pods_in_deployment(self.core_api, self.cluster_info.app_name)
         if len(pods) > 1:
             print("Warning: more than one pod in the deployment")
+        if len(pods) == 0:
+            log_data = "******* Pods not running ********\n"
+        else:
             k8s_pod_name = pods[0]
+            # If the pod is not yet started, the logs request below will throw an exception
+            try:
                 log_data = self.core_api.read_namespaced_pod_log(k8s_pod_name, namespace="default", container="test")
+            except client.exceptions.ApiException as e:
+                if opts.o.debug:
+                    print(f"Error from read_namespaced_pod_log: {e}")
+                log_data = "******* No logs available ********\n"
         return log_stream_from_string(log_data)
 
     def update(self):

stack_orchestrator/deploy/k8s/helpers.py

@@ -21,6 +21,7 @@ from typing import Set, Mapping, List
 
 from stack_orchestrator.opts import opts
 from stack_orchestrator.deploy.deploy_util import parsed_pod_files_map_from_file_names
+from stack_orchestrator.deploy.deployer import DeployerException
 
 
 def _run_command(command: str):
@@ -29,10 +30,13 @@ def _run_command(command: str):
     result = subprocess.run(command, shell=True)
     if opts.o.debug:
         print(f"Result: {result}")
+    return result
 
 
 def create_cluster(name: str, config_file: str):
-    _run_command(f"kind create cluster --name {name} --config {config_file}")
+    result = _run_command(f"kind create cluster --name {name} --config {config_file}")
+    if result.returncode != 0:
+        raise DeployerException(f"kind create cluster failed: {result}")
 
 
 def destroy_cluster(name: str):
@@ -41,12 +45,14 @@ def destroy_cluster(name: str):
 
 def load_images_into_kind(kind_cluster_name: str, image_set: Set[str]):
     for image in image_set:
-        _run_command(f"kind load docker-image {image} --name {kind_cluster_name}")
+        result = _run_command(f"kind load docker-image {image} --name {kind_cluster_name}")
+        if result.returncode != 0:
+            raise DeployerException(f"kind create cluster failed: {result}")
 
 
 def pods_in_deployment(core_api: client.CoreV1Api, deployment_name: str):
     pods = []
-    pod_response = core_api.list_namespaced_pod(namespace="default", label_selector="app=test-app")
+    pod_response = core_api.list_namespaced_pod(namespace="default", label_selector=f"app={deployment_name}")
     if opts.o.debug:
         print(f"pod_response: {pod_response}")
     for pod_info in pod_response.items:

tests/k8s-deploy/run-deploy-test.sh

@@ -2,13 +2,58 @@
 set -e
 if [ -n "$CERC_SCRIPT_DEBUG" ]; then
     set -x
-fi
-# Note: eventually this test should be folded into ../deploy/
-# but keeping it separate for now for convenience
-TEST_TARGET_SO=$( ls -t1 ./package/laconic-so* | head -1 )
     # Dump environment variables for debugging
     echo "Environment variables:"
     env
+fi
+
+# Helper functions: TODO move into a separate file
+wait_for_pods_started () {
+    for i in {1..5}
+    do
+        local ps_output=$( $TEST_TARGET_SO deployment --dir $test_deployment_dir ps )
+
+        if [[ "$ps_output" == *"Running containers:"* ]]; then
+            # if ready, return
+            return
+        else
+            # if not ready, wait
+            sleep 5
+        fi
+    done
+    # Timed out, error exit
+    echo "waiting for pods to start: FAILED"
+    delete_cluster_exit
+}
+
+wait_for_log_output () {
+    for i in {1..5}
+    do
+
+        local log_output=$( $TEST_TARGET_SO deployment --dir $test_deployment_dir logs )
+
+        if [[ ! -z "$log_output" ]]; then
+            # if ready, return
+            return
+        else
+            # if not ready, wait
+            sleep 5
+        fi
+    done
+    # Timed out, error exit
+    echo "waiting for pods log content: FAILED"
+    delete_cluster_exit
+}
+
+
+delete_cluster_exit () {
+    $TEST_TARGET_SO deployment --dir $test_deployment_dir stop --delete-volumes
+    exit 1
+}
+
+# Note: eventually this test should be folded into ../deploy/
+# but keeping it separate for now for convenience
+TEST_TARGET_SO=$( ls -t1 ./package/laconic-so* | head -1 )
 # Set a non-default repo dir
 export CERC_REPO_BASE_DIR=~/stack-orchestrator-test/repo-base-dir
 echo "Testing this package: $TEST_TARGET_SO"
@@ -53,23 +98,36 @@ fi
 echo "deploy create output file test: passed"
 # Try to start the deployment
 $TEST_TARGET_SO deployment --dir $test_deployment_dir start
-# TODO: add a check to see if the container is up
+wait_for_pods_started
-# Sleep because k8s not up yet
-sleep 30
 # Check logs command works
+wait_for_log_output
 log_output_3=$( $TEST_TARGET_SO deployment --dir $test_deployment_dir logs )
 if [[ "$log_output_3" == *"Filesystem is fresh"* ]]; then
     echo "deployment logs test: passed"
 else
     echo "deployment logs test: FAILED"
-    exit 1
+    delete_cluster_exit
 fi
 # Check the config variable CERC_TEST_PARAM_1 was passed correctly
 if [[ "$log_output_3" == *"Test-param-1: PASSED"* ]]; then
     echo "deployment config test: passed"
 else
     echo "deployment config test: FAILED"
-    exit 1
+    delete_cluster_exit
+fi
+# Stop then start again and check the volume was preserved
+$TEST_TARGET_SO deployment --dir $test_deployment_dir stop
+# Sleep a bit just in case
+sleep 2
+$TEST_TARGET_SO deployment --dir $test_deployment_dir start
+wait_for_pods_started
+wait_for_log_output
+log_output_4=$( $TEST_TARGET_SO deployment --dir $test_deployment_dir logs )
+if [[ "$log_output_4" == *"Filesystem is old"* ]]; then
+    echo "Retain volumes test: passed"
+else
+    echo "Retain volumes test: FAILED"
+    delete_cluster_exit
 fi
 # Stop and clean up
 $TEST_TARGET_SO deployment --dir $test_deployment_dir stop --delete-volumes