Open Metrics Framework

The Open Metrics Framework is an end-to-end analytics platform designed to capture, analyze, and convey the real-world performance of EBRAINS. It provides comprehensive KPI tracking aligned with project objectives, insights to address complex questions, and data-driven decision support. Furthermore, the framework enhances monitoring and observability across all EBRAINS services and infrastructure components, empowering owners to ensure their services remain reliable, scalable, and high-performing

Key Features of the Framework

Data Collection: It collects data across multiple categories:
- User Engagement
- Search And Download Activity
- Publications And Citations
- Support Tickets
- System Availability And Performance
Unified Monitoring: Provides seamless data visualization with both platform-wide and component-specific views, enabling teams to explore and analyze metrics efficiently.
Actionable Alerts: Offers real-time incident alerts ensure proactive monitoring, enabling teams to quickly identify and resolve potential issues to maintain platform reliability.

Technical Stack

The Open Metrics Framework Stack selection is based on two criteria:

Open Source status to eliminate licensing constraints
Capability for a single front-end to visualize data from diverse back-end types

It consits of:

Prometheus: For time series system metrics
OpenSearch: For data requiring complex analysis and aggregations, collected from external APIs (e.g. Matomo API)
Grafana: Provides a user-friendly interface with dashboards for tracking Key Performance Indicators (KPIs) and metrics
Kuma: For monitoring the health, uptime and response times of critical EBRAINS services (e.g. Drive, JupyterHub, etc.)