Supercomputer Dashboard

A comprehensive monitoring and management dashboard for SahasraT, the flagship supercomputer at the Supercomputer Education and Research Centre (SERC), Indian Institute of Science (IISc).

About SahasraT

SahasraT was India's fastest supercomputer during my tenure, featuring:

• 33,000+ CPU cores across multiple node types
• Petascale storage with Lustre parallel filesystem
• InfiniBand interconnect for high-speed node communication
• Serving 500+ researchers across India

Dashboard Features

Real-time Monitoring

• Node health status with color-coded visualization
• CPU, memory, and GPU utilization metrics
• Storage capacity and I/O performance
• Network traffic and InfiniBand statistics

Job Management

• PBS job queue visualization
• Historical job analytics and reporting
• Resource allocation trends
• User quota management

Alerting System

• Configurable threshold-based alerts
• Email and SMS notifications via Twilio
• Integration with Nagios for infrastructure alerts
• Incident tracking and resolution workflow

Technical Implementation

• Backend: Django 3.x with Django REST Framework
• Frontend: Bootstrap 4 with Plotly for visualizations
• Data Collection: Custom Python agents using PBS and IPMI APIs
• Database: PostgreSQL with TimescaleDB for time-series metrics
• Caching: Redis for real-time metric caching
• Analytics: Pandas for data processing and analysis