Helm tests for HA deployment (#41)

.github/workflows/lint-test-high-availability.yaml

@@ -1,4 +1,4 @@
-name: Lint and Test Charts
+name: Lint and Test Charts High Availability
 
 on:
   pull_request:
@@ -39,7 +39,6 @@ jobs:
       - name: Run chart-testing (lint)
         if: steps.list-changed.outputs.changed == 'true'
         run: ct lint --target-branch ${{ github.event.repository.default_branch }}
-        continue-on-error: true
 
       - name: Create kind cluster
         if: steps.list-changed.outputs.changed == 'true'
@@ -77,5 +76,4 @@ jobs:
       - name: Run Helm tests
         if: steps.list-changed.outputs.changed == 'true'
         run: |
-            helm test mem-ha-test
-            kubectl logs -l app=memgraph
+            timeout 30 helm test mem-ha-test  # If connection to some port is broken, this will timeout with code 124

charts/memgraph-high-availability/Chart.yaml

@@ -2,7 +2,7 @@ apiVersion: v2
 name: memgraph-high-availability
 description: A Helm chart for Kubernetes with Memgraph High availabiliy capabilites
 
-version: 0.1.1
+version: 0.1.2
 appVersion: "2.18.0"
 
 type: application

charts/memgraph-high-availability/templates/tests/test-cluster-setup.yaml

@@ -10,44 +10,47 @@ spec:
   template:
     metadata:
       labels:
-        app: memgraph
+        app: memgraph-ha-test-cluster-setup
     spec:
       containers:
       - name: memgraph-test-cluster-setup
-        image: memgraph/memgraph:2.17.0
+        image: "{{ $.Values.memgraph.image.repository }}:{{ $.Values.memgraph.image.tag }}"
         command: ["/bin/bash", "-c"]
         args:
         - |
-          echo "Running cluster test setup..."
-          result=$(echo 'SHOW INSTANCES;' | mgconsole --host memgraph-coordinator-1.default.svc.cluster.local --port 7687)
-          echo "$result"
+          instances=$(echo 'SHOW INSTANCES;' | mgconsole --host memgraph-coordinator-1.default.svc.cluster.local --port 7687)
+          echo "$instances"
+          num_leaders=$(echo "$instances" | grep -c "leader")
+          if [ $num_leaders -ne 1 ]; then
+              echo "Expected 1 leader after registration, got $num_leaders."
+              exit 1
+          fi
+
+          num_followers=$(echo "$instances" | grep -c "follower")
+          if [ $num_followers -ne 2 ]; then
+              echo "Expected 2 followers after registration, got $num_followers."
+              exit 1
+          fi
+
+          num_mains=$(echo "$instances" | grep -c "main")
+          if [ $num_mains -ne 1 ]; then
+              echo "Expected 1 main after registration, got $num_mains."
+              exit 1
+          fi
+
+          num_replicas=$(echo "$instances" | grep -c "replica")
+          if [ $num_replicas -ne 1 ]; then
+              echo "Expected 1 replica after registration, got $num_replicas."
+              exit 1
+          fi
+
+          num_instances_down=$(echo "instances" | grep -c "down")
+          if [ $num_instances_down -ne 0 ]; then
+              echo "Expected 0 instances down after registration, got $num_instances_down."
+              exit 1
+          fi
+
+          exit 0
 
-          # Use awk to check if all instances have health status "up" and there are exactly 5 instances
-          echo "$result" | awk '
-          BEGIN {
-              FS = "|"
-              instance_count = 0
-              health_ok = 1
-          }
-          NR > 3 && NR <= 8 {
-              gsub(/^ *| *$/, "", $6)  # Trim spaces from health
-              health = $6
-              print "Health: [" health "]"  # Debug print
-              instance_count++
-              if (health != "up") {
-                  health_ok = 0
-                  print "Non-up health found: [" health "]"  # Debug print
-              }
-          }
-          END {
-              print "Instance count: " instance_count  # Debug print
-              print "Health OK: " health_ok  # Debug print
-              if (instance_count == 5 && health_ok == 1) {
-                  print "All instances are up and there are exactly 5 instances."
-              } else {
-                  print "Check failed. Either not all instances are up or the count is not 5."
-              }
-          }
-          '
       restartPolicy: Never
   backoffLimit: 3

charts/memgraph-high-availability/templates/tests/test-connection.yaml

@@ -1,3 +1,4 @@
+# Inspect logs of this test with kubectl logs -l app=memgraph-ha-test-connection
 apiVersion: batch/v1
 kind: Job
 metadata:
@@ -8,22 +9,31 @@ metadata:
     "helm.sh/hook": test
 spec:
   template:
+    metadata:
+      labels:
+        app: memgraph-ha-test-connection
     spec:
       containers:
       - name: memgraph-test
-        image: memgraph/memgraph:2.17.0
-        command: ["/bin/sh", "-c"]
+        image: "{{ $.Values.memgraph.image.repository }}:{{ $.Values.memgraph.image.tag }}"
+        command: ["/bin/bash", "-c"]
         args:
           - |
-            echo "Running connection test on data instances..."
-            echo "RETURN 0;" | mgconsole --host memgraph-data-0.default.svc.cluster.local --port 7687
-            echo "RETURN 0;" | mgconsole --host memgraph-data-1.default.svc.cluster.local --port 7687
-
-            echo "Running a connection test on coordinator instances..."
-            echo "SHOW INSTANCES;" | mgconsole --host memgraph-coordinator-1.default.svc.cluster.local --port 7687
-            echo "SHOW INSTANCES;" | mgconsole --host memgraph-coordinator-2.default.svc.cluster.local --port 7687
-            echo "SHOW INSTANCES;" | mgconsole --host memgraph-coordinator-3.default.svc.cluster.local --port 7687
-
+            commands=(
+              "echo 'SHOW INSTANCES;' | mgconsole --host memgraph-coordinator-1.default.svc.cluster.local --port 7687"
+              "echo 'SHOW INSTANCES;' | mgconsole --host memgraph-coordinator-2.default.svc.cluster.local --port 7687"
+              "echo 'SHOW INSTANCES;' | mgconsole --host memgraph-coordinator-3.default.svc.cluster.local --port 7687"
+              "echo 'RETURN 0;' | mgconsole --host memgraph-data-0.default.svc.cluster.local --port 7687"
+              "echo 'RETURN 0;' | mgconsole --host memgraph-data-1.default.svc.cluster.local --port 7687"
+            )
+            for command in "${commands[@]}"; do
+              echo "Running command: $command"
+              eval $command
+              if [ $? -ne 0 ]; then
+                echo "Command failed: $command"
+                exit 1
+              fi
+            done
+            exit 0
 
       restartPolicy: Never
-  backoffLimit: 3

charts/memgraph-high-availability/values.yaml

@@ -38,7 +38,6 @@ memgraph:
       logPVCSize: "256Mi"
 
 
-
 data:
 - id: "0"
   boltPort: 7687