stefanprodan/dockprom
Dockprom
dockprom is a monitoring stack based on Prometheus and Grafana designed to track the performance of Docker containers and their underlying hosts. It functions as a complete solution for gathering real-time metrics and displaying them through a self-hosted dashboard.
The project includes a suite of tools for collecting container and host metrics, as well as a discovery tool specifically for automatically identifying and adding tagged EC2 instances to the monitoring configuration.
The system covers several observability areas, including time-series data storage and the creation of performance dashboards for CPU, memory, storage, and network activity. It also features a centralized alerting system to route notifications when resource thresholds are exceeded and a reverse-proxy to manage access control and security for the monitoring dashboards.
The stack is deployed as a collection of interdependent services using Docker Compose.
Features
- Resource Monitoring - Provides a complete suite for tracking the performance and resource usage of Docker containers and hosts.
- Time Series Data Storage - Utilizes time-series data storage to maintain historical records of numerical performance metrics.
- Docker Compose Deployments - Deploys the entire monitoring stack as a collection of interdependent services using Docker Compose.
- Container Metrics Collectors - Bundles tools like cAdvisor and NodeExporter to gather real-time metrics from Docker environments.
- Pull-Based Metric Scraping - Gathers system and container performance data by periodically scraping metrics from exposed HTTP endpoints.
- System Metrics Collection - Gathers real-time performance metrics from both Docker containers and the underlying host machine.
- Performance Visualization - Creates real-time dashboards to visualize CPU, memory, and network activity across distributed servers.
- Performance Dashboards - Provides interactive dashboards for real-time monitoring of CPU, memory, storage, and network activity.
- Prometheus Monitoring Practices - Ships a complete monitoring stack based on Prometheus and Grafana for system observability and analysis.
- Time-Series Monitoring Dashboards - Transforms raw time-series data into graphical dashboards for real-time monitoring of system metrics.
- Self-Hosted System Dashboards - Provides a self-hosted web interface for monitoring CPU, memory, and network activity across servers.
- Cloud Resource Discovery - Automatically identifies and adds EC2 instances to the monitoring system based on cloud metadata tags.
- Service Discovery Integrations - Integrates service discovery to automatically identify and add cloud host targets to the monitoring list.
- Host Auto-Discovery - Automatically identifies tagged EC2 cloud instances and adds them to the monitoring list.
- Alert Notification Routing - Routes system alerts to external communication channels via a centralized manager based on resource thresholds.
- Alert Routing - Provides a centralized manager to filter and forward triggered alerts to external communication channels.
- Performance Degradation Alerts - Defines rules and thresholds for CPU, memory, and storage to trigger notifications during performance degradation.
- Alerting Systems - Configures AlertManager to route system notifications when resource usage exceeds defined thresholds.
- Alert Management Systems - Implements automated notifications and routing rules when server metrics exceed predefined performance thresholds.
- Observability and Monitoring - Full monitoring stack with Prometheus and Grafana.
