solisamicus opened a new pull request, #3047:
URL: https://github.com/apache/dubbo-go/pull/3047
feat: add HTTP health check API for Kubernetes integration
Add comprehensive HTTP health check support for liveness and readiness
probes to enable
seamless Kubernetes deployment and container orchestration.
## Background
Kubernetes requires HTTP endpoints for health probes to determine container
lifecycle:
- Liveness probes: detect if container should be restarted
- Readiness probes: control traffic routing to healthy instances
This implementation provides a robust, extensible health check system that
integrates
with dubbo-go's existing architecture while allowing users to define custom
health logic.
## Implementation
### Core Components
- **HealthCheckConfig**: YAML-configurable settings for ports, paths, and
timeouts
- **HealthChecker Interface**: Extensible interface for user-defined health
logic
- **HealthCheckServer**: HTTP server exposing /health/live and /health/ready
endpoints
- **Built-in Checkers**: Ready-to-use implementations for common scenarios
- **Global Registry**: Thread-safe system for managing health checkers
### Key Features
- **Configurable Endpoints**: Customizable ports and paths via configuration
- **Extensible Logic**: Users can implement HealthChecker interface for
custom checks
- **Composite Support**: Combine multiple health checkers with AND logic
- **Detailed Responses**: JSON responses with status, timestamp, and
detailed information
- **Timeout Protection**: Configurable timeouts prevent hanging health checks
- **Graceful Integration**: Seamless integration with dubbo-go lifecycle
management
- **Thread Safety**: Concurrent-safe health checker registry
### Changes Made
- Add HealthCheckConfig to MetricsConfig with port, paths, timeout settings
- Implement HealthChecker interface for user-defined health check logic
- Add HealthCheckServer with /health/live and /health/ready endpoints
- Support CompositeHealthChecker for combining multiple checkers
- Provide built-in checkers: Default, Dubbo, GracefulShutdown, Timeout
- Integrate with metrics system for automatic server lifecycle management
- Add thread-safe global health checker registry
- Support detailed health results with JSON response format
- Include graceful shutdown integration and timeout protection
## Usage
### Configuration
```yaml
metrics:
enable: true
health-check:
enabled: true
port: "8080"
live-path: "/health/live"
ready-path: "/health/ready"
timeout: "10s"
```
### Custom Health Checker
```go
type MyHealthChecker struct {
db *sql.DB
}
func (m *MyHealthChecker) CheckLiveness(ctx context.Context) bool {
return true // Process is alive
}
func (m *MyHealthChecker) CheckReadiness(ctx context.Context) bool {
return m.db.Ping() == nil // Check database connection
}
func (m *MyHealthChecker) Name() string {
return "MyApp"
}
// Register custom checker
server.SetHealthChecker(&MyHealthChecker{db: database})
```
### Kubernetes Deployment
```yaml
livenessProbe:
httpGet:
path: /health/live
port: 8080
initialDelaySeconds: 30
periodSeconds: 10
failureThreshold: 3
readinessProbe:
httpGet:
path: /health/ready
port: 8080
initialDelaySeconds: 10
periodSeconds: 5
failureThreshold: 3
```
## API Responses
### Healthy Response (200 OK)
```json
{
"status": "UP",
"timestamp": 1640995200000,
"details": {
"check": "readiness",
"database": "connected",
"services": "exported"
}
}
```
### Unhealthy Response (503 Service Unavailable)
```json
{
"status": "DOWN",
"timestamp": 1640995200000,
"message": "Database connection failed",
"details": {
"check": "readiness",
"database": "unavailable",
"reason": "connection_timeout"
}
}
```
## Benefits
- **Cloud Native**: Full Kubernetes compatibility with standard probe
endpoints
- **Zero Downtime**: Proper readiness checks enable rolling deployments
- **Fault Tolerance**: Automatic unhealthy instance removal from load
balancers
- **Observability**: Detailed health status information for debugging
- **Extensibility**: Plugin architecture for custom health logic
- **Production Ready**: Timeout protection, graceful shutdown, error handling
- **Backward Compatible**: Disabled by default, no impact on existing
deployments
## Testing
Health check endpoints can be tested using:
```bash
curl http://localhost:8080/health/live
curl http://localhost:8080/health/ready
```
Fixes: https://github.com/apache/dubbo-go/issues/2039
Enables: Kubernetes-native health monitoring and container lifecycle
management
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
