[deploy] add Helm chart #182
Conversation
WalkthroughAdds a new Helm chart for sms-gateway (Chart, templates, values, README, NOTES, tests), a Minikube test script, and a GitHub Actions workflow to package and push the chart to an S3-backed Helm repo (package + push; index update not included). Changes
Sequence Diagram(s)sequenceDiagram
autonumber
actor Dev as Developer
participant GH as GitHub Actions
participant Helm as Helm CLI
participant S3 as AWS S3 (helm-s3 plugin)
Dev->>GH: Push to master (deployments/helm-chart) or create release
GH->>GH: Checkout repo
GH->>Helm: Setup Helm + install helm-s3 plugin
GH->>Helm: helm package deployments/helm-chart (tag or 0.0.0-dev)
GH->>S3: helm s3 init / helm s3 add repo (if missing)
GH->>S3: helm s3 push <chart.tgz> s3://bucket/repo
GH-->>Dev: Workflow completes (artifacts cleaned)
sequenceDiagram
autonumber
actor Operator as Cluster Operator
participant Helm as Helm CLI
participant K8s as Kubernetes API
participant App as sms-gateway
participant DB as MariaDB
Operator->>Helm: helm install/upgrade sms-gateway -f values.yaml
Helm->>K8s: Apply manifests (Deployment, Service, ...)
alt database.deployInternal = true
K8s->>K8s: Create StatefulSet/Service/PVC/Secret for MariaDB
App->>DB: Connect via internal service
else
App->>DB: Connect to external DB host/port
end
K8s->>App: Pods become Ready
Operator->>App: Access via Ingress / Service / NodePort / LoadBalancer
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Pre-merge checks and finishing touches✅ Passed checks (3 passed)
✨ Finishing touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 8
🧹 Nitpick comments (8)
deployments/helm-chart/README.md (1)
54-57: Align docs with actual values keyThe chart templates expect
gateway.fcmCredentials, but the table documentsgateway.fcmKey. Please update the docs (or the values) so the key matches reality; otherwise users will set the wrong field.deployments/helm-chart/values.yaml (1)
73-73: Database host should reference the internal service name.When
database.deployInternalis true, the database host should match the generated service name pattern (i.e.,<fullname>-db) rather than a staticdbvalue. The deployment template correctly handles this at lines 44-50 of deployment.yaml, but the default value here is misleading.Consider updating the default or adding a comment:
database: - host: db + host: "" # Auto-configured when deployInternal is true, otherwise set your external DB host port: 3306deployments/helm-chart/templates/database.yaml (2)
46-57: Consider adding initial delay to probes.The readiness and liveness probes start immediately, which might cause unnecessary failures during MariaDB initialization, especially for the first startup.
Add initial delays:
readinessProbe: + initialDelaySeconds: 30 + periodSeconds: 10 exec: command: - healthcheck.sh - --connect - --innodb_initialized livenessProbe: + initialDelaySeconds: 60 + periodSeconds: 30 exec: command: - healthcheck.sh - --connect - --innodb_initialized
102-106: Consider making storage class configurable.The PersistentVolumeClaim doesn't specify a
storageClassName, which means it will use the cluster's default storage class. This may not be suitable for all environments.Add a configurable storage class in values.yaml and reference it here:
In values.yaml:
database: mariadb: persistence: enabled: true size: 8Gi storageClass: "" # Use default if emptyIn database.yaml:
resources: requests: storage: {{ .Values.database.mariadb.persistence.size }} + {{- if .Values.database.mariadb.persistence.storageClass }} + storageClassName: {{ .Values.database.mariadb.persistence.storageClass }} + {{- end }} {{- end }}deployments/helm-chart/templates/deployment.yaml (2)
83-90: Add initial delays and timeouts to health probes.The health probes lack initial delay and timeout configurations, which could cause premature failures during application startup or when the application is under load.
Add probe configuration:
livenessProbe: + initialDelaySeconds: 30 + periodSeconds: 10 + timeoutSeconds: 5 + failureThreshold: 3 httpGet: path: /health port: http readinessProbe: + initialDelaySeconds: 10 + periodSeconds: 5 + timeoutSeconds: 3 + failureThreshold: 3 httpGet: path: /health port: http
60-76: Verify mutual exclusivity of gateway modes.The template assumes that either
privateTokenorfcmCredentialsis set, but doesn't handle the case where both are set or neither is set. This could lead to an incomplete deployment configuration.Consider adding validation or documentation:
- Add a comment explaining the modes:
{{- if .Values.gateway.privateToken }} # Private gateway mode - uses private token for authentication - name: GATEWAY__MODE
- Consider adding a validation check that fails deployment if the configuration is invalid. Would you like me to help generate a validation helper that checks these conditions?
scripts/test-helm-minikube.sh (2)
66-69: Consider less aggressive cleanup option.Deleting the entire Minikube cluster (
minikube delete) is aggressive and will destroy all resources, not just the test namespace. This could be problematic if the user has other workloads in the cluster.Consider only deleting the namespace:
# Cleanup echo "Cleaning up..." kill $PORT_FORWARD_PID -minikube delete +kubectl delete namespace $NAMESPACE +echo "Note: Minikube cluster is still running. Run 'minikube delete' to remove it."Or make it optional:
# Cleanup echo "Cleaning up..." kill $PORT_FORWARD_PID + +# Delete namespace +kubectl delete namespace $NAMESPACE + +# Optionally delete Minikube cluster +read -p "Delete Minikube cluster? (y/N): " -n 1 -r +echo +if [[ $REPLY =~ ^[Yy]$ ]]; then + minikube delete +fi
50-56: Improve reliability of port-forward setup.The script uses a fixed 5-second sleep to wait for port-forward, which may be insufficient on slower systems or excessive on faster ones.
Consider actively waiting for the port:
# Port forward to access the service echo "Port forwarding to service (http://localhost:8080)..." kubectl port-forward --namespace $NAMESPACE service/sms-gateway-test 8080:3000 & PORT_FORWARD_PID=$! -# Give it a moment to establish the connection -sleep 5 +# Wait for port-forward to be ready +echo "Waiting for port-forward to be ready..." +for i in {1..30}; do + if curl -s http://localhost:8080/health > /dev/null 2>&1; then + echo "Port-forward is ready!" + break + fi + if [ $i -eq 30 ]; then + echo "Error: Port-forward failed to establish" + kill $PORT_FORWARD_PID + exit 1 + fi + sleep 1 +done
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (15)
.github/workflows/publish-helm-chart.yml(1 hunks)deployments/helm-chart/Chart.yaml(1 hunks)deployments/helm-chart/README.md(1 hunks)deployments/helm-chart/templates/NOTES.txt(1 hunks)deployments/helm-chart/templates/_helpers.tpl(1 hunks)deployments/helm-chart/templates/database.yaml(1 hunks)deployments/helm-chart/templates/deployment.yaml(1 hunks)deployments/helm-chart/templates/hpa.yaml(1 hunks)deployments/helm-chart/templates/ingress.yaml(1 hunks)deployments/helm-chart/templates/secrets.yaml(1 hunks)deployments/helm-chart/templates/service.yaml(1 hunks)deployments/helm-chart/templates/serviceaccount.yaml(1 hunks)deployments/helm-chart/templates/tests/test-connection.yaml(1 hunks)deployments/helm-chart/values.yaml(1 hunks)scripts/test-helm-minikube.sh(1 hunks)
🧰 Additional context used
🧠 Learnings (1)
📚 Learning: 2025-09-05T08:37:28.077Z
Learnt from: eznix86
PR: android-sms-gateway/server#172
File: deployments/kubernetes/sms-gateway-server/Chart.yaml:9-10
Timestamp: 2025-09-05T08:37:28.077Z
Learning: In the android-sms-gateway/server project, the Helm chart's appVersion in Chart.yaml is set to "latest" as a placeholder because CI automatically updates it with the actual version from git tags during the release process via the bump-chart-app-version.yaml workflow.
Applied to files:
deployments/helm-chart/README.mddeployments/helm-chart/Chart.yaml
🪛 Shellcheck (0.11.0)
scripts/test-helm-minikube.sh
[warning] 27-27: Quote this to prevent word splitting.
(SC2046)
🪛 YAMLlint (1.37.1)
deployments/helm-chart/templates/service.yaml
[error] 6-6: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/hpa.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/secrets.yaml
[error] 4-4: syntax error: expected , but found ''
(syntax)
deployments/helm-chart/templates/ingress.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/database.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/serviceaccount.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/tests/test-connection.yaml
[error] 4-4: syntax error: expected , but found ''
(syntax)
deployments/helm-chart/templates/deployment.yaml
[error] 6-6: syntax error: expected the node content, but found '-'
(syntax)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (2)
- GitHub Check: E2E
- GitHub Check: Analyze (go)
🔇 Additional comments (7)
deployments/helm-chart/values.yaml (1)
1-99: LGTM overall structure!The values.yaml structure is well-organized with comprehensive configuration options for deployment, service, ingress, resources, autoscaling, database, and gateway settings. The defaults are reasonable for development/testing scenarios.
deployments/helm-chart/templates/hpa.yaml (1)
1-28: LGTM!The HorizontalPodAutoscaler template is correctly configured with:
- Proper conditional rendering based on
.Values.autoscaling.enabled- autoscaling/v2 API version
- Both CPU and memory utilization metrics
- Correct reference to the target deployment
Note: The YAMLlint syntax error is a false positive caused by Helm template syntax (the
{{-delimiter).deployments/helm-chart/templates/NOTES.txt (1)
1-33: Well-structured post-deployment guidance.The NOTES.txt provides comprehensive instructions for accessing the application across different service types and includes helpful health check and database information.
deployments/helm-chart/templates/database.yaml (1)
1-107: Well-structured database deployment template.The database.yaml correctly configures a StatefulSet with:
- Proper secret references for passwords
- Health checks using MariaDB's built-in healthcheck script
- Persistent volume for data
- Appropriate labels and selectors
deployments/helm-chart/templates/deployment.yaml (1)
1-102: Excellent deployment template structure!The deployment.yaml is well-configured with:
- Proper environment variable handling for both internal and external database modes
- Conditional gateway mode configuration
- Correct secret references
- Resource limits and security contexts
- Comprehensive pod scheduling options (nodeSelector, affinity, tolerations)
scripts/test-helm-minikube.sh (1)
1-72: Well-structured test automation script!The script provides comprehensive end-to-end testing with proper error handling, prerequisite checks, and clear progress output. The overall flow is logical and helpful for testing the Helm chart.
deployments/helm-chart/templates/_helpers.tpl (1)
1-42: LGTM! Standard Helm helpers implementation.The helper templates follow Helm best practices with:
- Proper fullname generation with override support and truncation
- Standard label templates for consistent resource labeling
- Service account name resolution logic
- Correct handling of chart versioning in labels
These helpers provide the foundation for consistent naming and labeling across all chart resources.
There was a problem hiding this comment.
Actionable comments posted: 1
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (8)
.github/workflows/publish-helm-chart.yml(1 hunks)deployments/helm-chart/Chart.yaml(1 hunks)deployments/helm-chart/README.md(1 hunks)deployments/helm-chart/templates/NOTES.txt(1 hunks)deployments/helm-chart/templates/database.yaml(1 hunks)deployments/helm-chart/templates/deployment.yaml(1 hunks)deployments/helm-chart/values.yaml(1 hunks)scripts/test-helm-minikube.sh(1 hunks)
✅ Files skipped from review due to trivial changes (1)
- deployments/helm-chart/README.md
🚧 Files skipped from review as they are similar to previous changes (3)
- deployments/helm-chart/values.yaml
- deployments/helm-chart/Chart.yaml
- scripts/test-helm-minikube.sh
🧰 Additional context used
🪛 YAMLlint (1.37.1)
deployments/helm-chart/templates/database.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/deployment.yaml
[error] 6-6: syntax error: expected the node content, but found '-'
(syntax)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (2)
- GitHub Check: E2E
- GitHub Check: Analyze (go)
capcom6
force-pushed
the
deploy/helm-chart
branch
from
b87ed07 to
6123028
Compare
There was a problem hiding this comment.
Actionable comments posted: 2
🧹 Nitpick comments (2)
deployments/helm-chart/templates/tests/test-connection.yaml (1)
7-14: Add a hook-delete policy for Helm tests.Without a
helm.sh/hook-delete-policyannotation, successful test Pods linger unless the operator passeshelm test --cleanup. Addbefore-hook-creation,hook-succeededso repeated test runs stay tidy.annotations: "helm.sh/hook": test + "helm.sh/hook-delete-policy": before-hook-creation,hook-succeededdeployments/helm-chart/templates/database.yaml (1)
62-113: Prefer StatefulSet volumeClaimTemplates over detached PVCs.Hard-wiring a single PVC with
claimNamecouples the StatefulSet to a pre-created volume and blocks any scale-out or re-creation without manual cleanup. Define the storage involumeClaimTemplatesso Kubernetes provisions per-pod PVCs automatically.- volumes: - - name: mariadb-data - persistentVolumeClaim: - claimName: {{ include "sms-gateway.fullname" . }}-db-pvc + volumeClaimTemplates: + - metadata: + name: mariadb-data + spec: + accessModes: + - ReadWriteOnce + resources: + requests: + storage: {{ .Values.database.mariadb.persistence.size }} + {{- if .Values.database.mariadb.persistence.storageClass }} + storageClassName: {{ .Values.database.mariadb.persistence.storageClass }} + {{- end }}And drop the standalone PVC manifest below.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (15)
.github/workflows/publish-helm-chart.yml(1 hunks)deployments/helm-chart/Chart.yaml(1 hunks)deployments/helm-chart/README.md(1 hunks)deployments/helm-chart/templates/NOTES.txt(1 hunks)deployments/helm-chart/templates/_helpers.tpl(1 hunks)deployments/helm-chart/templates/database.yaml(1 hunks)deployments/helm-chart/templates/deployment.yaml(1 hunks)deployments/helm-chart/templates/hpa.yaml(1 hunks)deployments/helm-chart/templates/ingress.yaml(1 hunks)deployments/helm-chart/templates/secrets.yaml(1 hunks)deployments/helm-chart/templates/service.yaml(1 hunks)deployments/helm-chart/templates/serviceaccount.yaml(1 hunks)deployments/helm-chart/templates/tests/test-connection.yaml(1 hunks)deployments/helm-chart/values.yaml(1 hunks)scripts/test-helm-minikube.sh(1 hunks)
🚧 Files skipped from review as they are similar to previous changes (6)
- deployments/helm-chart/values.yaml
- deployments/helm-chart/README.md
- scripts/test-helm-minikube.sh
- deployments/helm-chart/templates/_helpers.tpl
- .github/workflows/publish-helm-chart.yml
- deployments/helm-chart/Chart.yaml
🧰 Additional context used
🪛 YAMLlint (1.37.1)
deployments/helm-chart/templates/secrets.yaml
[error] 4-4: syntax error: expected , but found ''
(syntax)
deployments/helm-chart/templates/tests/test-connection.yaml
[error] 4-4: syntax error: expected , but found ''
(syntax)
deployments/helm-chart/templates/hpa.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/database.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/deployment.yaml
[error] 6-6: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/ingress.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/service.yaml
[error] 6-6: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/serviceaccount.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (2)
- GitHub Check: E2E
- GitHub Check: Analyze (go)
There was a problem hiding this comment.
Actionable comments posted: 1
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (3)
deployments/helm-chart/templates/NOTES.txt(1 hunks)deployments/helm-chart/templates/ingress.yaml(1 hunks)deployments/helm-chart/templates/tests/test-connection.yaml(1 hunks)
🧰 Additional context used
🪛 YAMLlint (1.37.1)
deployments/helm-chart/templates/ingress.yaml
[error] 1-1: syntax error: expected the node content, but found '-'
(syntax)
deployments/helm-chart/templates/tests/test-connection.yaml
[error] 4-4: syntax error: expected , but found ''
(syntax)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (2)
- GitHub Check: E2E
- GitHub Check: Analyze (go)
| spec: | ||
| containers: | ||
| - name: curl | ||
| image: curlimages/curl:latest |
There was a problem hiding this comment.
Pin the curl test image for reproducibility and supply-chain safety.
curlimages/curl:latest drifts over time, so CI runs can pull different images with breaking changes or unpatched CVEs. Please lock it to a specific tag (ideally with a digest) and bump intentionally later.
- image: curlimages/curl:latest
+ image: curlimages/curl:8.11.0📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| image: curlimages/curl:latest | |
| image: curlimages/curl:8.11.0 |
🤖 Prompt for AI Agents
In deployments/helm-chart/templates/tests/test-connection.yaml around line 13,
the test pod image uses the floating tag "curlimages/curl:latest"; replace it
with a pinned image tag or digest (e.g., curlimages/curl:<specific-version> or
curlimages/curl@sha256:<digest>) to ensure reproducible CI runs and supply-chain
safety, updating the chart values or template accordingly and documenting the
chosen version so it can be intentionally bumped later.
1 participant
Summary by CodeRabbit