OpenShift OVN-Kubernetes Benchmark MCP Server

# OVN Kubernetes Container Usage Monitoring Project ## Project Structure ``` project_root/ ├── ovnk_benchmark_auth.py # Authentication module (provided) ├── ovnk_benchmark_prometheus_basequery.py # Base Prometheus query module (provided) ├── metrics-base.yml # Metrics configuration file (optional) ├── tools/ │ └── ovnk_benchmark_prometheus_ovnk_containers.py # Main collector module └── analysis/ └── ovnk_benchmark_performance_analysis_ovnk_containers.py # Performance analysis module ``` ## Files Created ### 1. `/tools/ovnk_benchmark_prometheus_ovnk_containers.py` **Class: `OVNKContainersUsageCollector`** Features: - **YAML Configuration Support**: Reads PromQL queries from `metrics-base.yml` file - **Fallback to Hardcoded Queries**: Uses built-in PromQL if YAML file is missing/invalid - Collects CPU, memory, and DB size metrics for OVN Kubernetes containers - Supports both instant and duration-based collection - Uses PrometheusBaseQuery for Prometheus integration - Integrates with OpenShiftAuth for cluster information - Exports results to JSON format - Concurrent query execution for performance Key Methods: - `__init__(prometheus_client, metrics_config_file)` - Initialize with optional YAML file path - `collect_instant_metrics()` - Collect metrics at a specific point in time - `collect_duration_metrics()` - Collect metrics over a time duration with min/max/avg - `reload_metrics_config()` - Dynamically reload configuration from YAML - `get_metrics_summary()` - Get summary of loaded configuration - `export_to_json()` - Export collected data to JSON file ### 2. `/analysis/ovnk_benchmark_performance_analysis_ovnk_containers.py` **Class: `OVNKPerformanceAnalyzer`** Features: - Analyzes performance data from the collector - Generates performance alerts with severity levels - Identifies top resource consumers - Calculates cluster health score (0-100) - Provides container-specific analysis - Generates human-readable performance reports - Configurable thresholds for different resource types ### 3. `metrics-base.yml` - Configuration File **Format**: YAML with metrics definitions Features: - **Flexible Configuration**: Define custom PromQL queries - **Automatic Fallback**: Uses hardcoded queries if file missing - **Dynamic Reloading**: Can be reloaded at runtime - **Rich Metadata**: Supports metricName, unit, description fields ## Metrics Configuration Priority 1. **YAML File** (`metrics-base.yml`): Primary source if file exists and is valid 2. **Hardcoded Queries**: Fallback if YAML file is missing, empty, or invalid 3. **Runtime Reload**: Configuration can be reloaded without restarting ## YAML Configuration Format ```yaml metrics: - query: 'your_promql_query_here' metricName: YOUR_METRIC_NAME unit: bytes|percentage|count|etc description: "# OVN Kubernetes Container Usage Monitoring Project ## Project Structure ``` project_root/ ├── ovnk_benchmark_auth.py # Authentication module (provided) ├── ovnk_benchmark_prometheus_basequery.py # Base Prometheus query module (provided) ├── tools/ │ └── ovnk_benchmark_prometheus_ovnk_containers.py # Main collector module └── analysis/ └── ovnk_benchmark_performance_analysis_ovnk_containers.py # Performance analysis module ``` ## Files Created ### 1. `/tools/ovnk_benchmark_prometheus_ovnk_containers.py` **Class: `OVNKContainersUsageCollector`** Features: - Collects CPU, memory, and DB size metrics for OVN Kubernetes containers - Supports both instant and duration-based collection - Uses PrometheusBaseQuery for Prometheus integration - Integrates with OpenShiftAuth for cluster information - Exports results to JSON format - Built-in metrics configuration with default PromQL queries - Concurrent query execution for performance Key Methods: - `collect_instant_metrics()` - Collect metrics at a specific point in time - `collect_duration_metrics()` - Collect metrics over a time duration with min/max/avg - `export_to_json()` - Export collected data to JSON file - `get_cluster_info()` - Get cluster information via OpenShiftAuth ### 2. `/analysis/ovnk_benchmark_performance_analysis_ovnk_containers.py` **Class: `OVNKPerformanceAnalyzer`** Features: - Analyzes performance data from the collector - Generates performance alerts with severity levels - Identifies top resource consumers - Calculates cluster health score (0-100) - Provides container-specific analysis - Generates human-readable performance reports - Configurable thresholds for different resource types Key Methods: - `analyze_performance_data()` - Main analysis function - `generate_performance_report()` - Human-readable report generation - `get_container_analysis()` - Detailed analysis for specific containers - `export_analysis()` - Export analysis results to JSON ## Default Metrics Configuration The collector includes these built-in PromQL queries: ### CPU Metrics - Container CPU usage percentage: `rate(container_cpu_usage_seconds_total{pod=~"ovnkube-master.*|ovnkube-node.*", container!="POD", container!=""}[5m]) * 100` ### Memory Metrics - Container memory RSS: `container_memory_rss{pod=~"ovnkube-master.*|ovnkube-node.*", container!="POD", container!=""}` - Container memory working set: `container_memory_working_set_bytes{pod=~"ovnkube-master.*|ovnkube-node.*", container!="POD", container!=""}` ### OVN Database Size Metrics - NBDB size: `ovn_db_db_size_bytes{db_name=~"OVN_Northbound"}` - SBDB size: `ovn_db_db_size_bytes{db_name=~"OVN_Southbound"}` ### OVN Container-Specific Memory Metrics - NBDB memory: `sum by(pod) (container_memory_rss{pod=~"ovnkube-master.*|ovnkube-node.*", container=~"nbdb"})` - SBDB memory: `sum by(pod) (container_memory_rss{pod=~"ovnkube-master.*|ovnkube-node.*", container=~"sbdb"})` - Northd memory: `sum by(pod) (container_memory_rss{pod=~"ovnkube-master.*|ovnkube-node.*", container=~"northd"})` - OVN Kube Controller memory: `sum by(pod) (container_memory_rss{pod=~"ovnkube-master.*|ovnkube-node.*", container=~"ovnkube-controller"})` - OVN Controller memory: `sum by(pod) (container_memory_rss{pod=~"ovnkube-master.*|ovnkube-node.*", container=~"ovn-controller"})` ## Usage Examples ### Basic Usage ```python import asyncio from ovnk_benchmark_prometheus_basequery import PrometheusBaseQuery from ovnk_benchmark_auth import auth from tools.ovnk_benchmark_prometheus_ovnk_containers import OVNKContainersUsageCollector from analysis.ovnk_benchmark_performance_analysis_ovnk_containers import OVNKPerformanceAnalyzer async def main(): # Initialize authentication await auth.initialize() # Create Prometheus client prometheus_client = PrometheusBaseQuery( prometheus_url=auth.prometheus_url, token=auth.prometheus_token ) # Create collector collector = OVNKContainersUsageCollector(prometheus_client) # Collect instant metrics instant_data = await collector.collect_instant_metrics() # Collect duration metrics (last 5 minutes) duration_data = await collector.collect_duration_metrics("5m") # Export to JSON instant_file = collector.export_to_json(instant_data) duration_file = collector.export_to_json(duration_data) # Analyze performance analyzer = OVNKPerformanceAnalyzer() analysis = analyzer.analyze_performance_data(duration_data) # Generate report report = analyzer.generate_performance_report(analysis) print(report) # Export analysis analysis_file = analyzer.export_analysis(analysis) await prometheus_client.close() if __name__ == "__main__": asyncio.run(main()) ``` ### Command Line Usage ```bash # Run the collector python tools/ovnk_benchmark_prometheus_ovnk_containers.py # Analyze existing metrics file python analysis/ovnk_benchmark_performance_analysis_ovnk_containers.py metrics_file.json ``` ## JSON Output Format ### Collector Output ```json { "collection_timestamp": "2025-08-25T10:30:00.000Z", "collection_type": "duration", "start_time": "2025-08-25T10:25:00.000Z", "end_time": "2025-08-25T10:30:00.000Z", "duration": "5m", "metrics": { "container_cpu_usage": { "config": {...}, "result_type": "duration", "data": [ { "pod_name": "ovnkube-master-xyz", "container_name": "nbdb", "min": 2.5, "max": 15.8, "avg": 8.2, "unit": "percentage", "sample_count": 20, "labels": {...} } ] } }, "summary": { "total_metrics_collected": 10, "successful_metrics": 9, "failed_metrics": 1, "top_usage": { "cpu": [...], "memory": [...], "db_size": [...] }, "containers_summary": {...}, "db_summary": {...} } } ``` ### Analysis Output ```json { "analysis_timestamp": "2025-08-25T10:31:00.000Z", "total_containers": 25, "total_pods": 8, "cluster_health_score": 87.5, "performance_alerts": [ { "severity": "high", "metric_type": "memory", "pod_name": "ovnkube-master-abc", "container_name": "nbdb", "message": "High memory usage detected", "current_value": 1536000000, "threshold": 1073741824, "unit": "bytes", "timestamp": "2025-08-25T10:31:00.000Z" } ], "top_performers": {...}, "resource_utilization": {...}, "recommendations": [ "📊 OVN Database containers showing high memory usage. Consider OVN database compaction or cleanup.", "⚖️ Resource imbalance detected in pods: ovnkube-node-xyz. Consider container resource limit tuning." ] } ``` ## Performance Thresholds ### CPU Usage - High: > 80% - Critical: > 95% ### Memory Usage - High: > 1GB - Critical: > 2GB ### Database Size - High: > 100MB - Critical: > 500MB ## Features 1. **Concurrent Metrics Collection** - All Prometheus queries run in parallel for efficiency 2. **Flexible Time Ranges** - Support for both instant and duration-based collection 3. **Comprehensive Analysis** - Performance alerts, health scoring, and recommendations 4. **OVN-Aware** - Special handling for OVN Kubernetes components 5. **Export Capabilities** - JSON export for both metrics and analysis 6. **Human-Readable Reports** - Formatted text reports for easy consumption 7. **Error Handling** - Graceful handling of failed queries and missing data 8. **Extensible Configuration** - Easy to add new metrics or modify thresholds