Skip to main content

Technical architecture

GridCapa is an open-source platform designed for capacity assessment in electricity networks. It is built around a modular, event-driven microservices architecture to support scalability, automation, and integration with industrial processes.

Modular Microservices Architecture​

GridCapa is composed of multiple loosely-coupled services, each with a specific role:

  • data-bridge — monitors FTP input
  • task-manager — orchestrates task lifecycle
  • process-runner — performs computation
  • rao-runner — solves optimization problems
  • gridcapa-app — user interface (HMI)
  • gridcapa-export — handles result exports

Event-Driven Workflow Orchestration​

  • Communication between services is handled via RabbitMQ.
  • Events (e.g. new file detected, task status change) trigger appropriate actions across the system.

Input/Output File Management​

  • Input files are uploaded to an FTP server and moved to MinIO object storage.
  • Output files and logs are exported back to the FTP server for external use.

Task Lifecycle Management​

Tasks go through several lifecycle stages:

  • READY – all inputs available
  • PENDING – waiting for a computation instance
  • RUNNING – actively computing
  • SUCCESS or ERROR – computation completed

Flexible Task Execution​

Tasks can be launched:

  • Automatically when all inputs are ready and config allows auto-start
  • Manually via the user interface

Computation and Adaptation Layer​

  • process-adapter formats task info and routes it to the process-runner
  • process-runner:
    • Preprocesses input data
    • Runs iterative logic (e.g. dichotomy)
    • Invokes rao-runner for optimization
    • Writes results to storage

Real-Time Monitoring and Logging​

  • Logs are continuously collected and stored in the database
  • UI updates in real time via WebSockets
  • Optimization logs are filtered by rao-logs-dispatcher

Output Handling and Export​

  • gridcapa-export copies result files and logs from MinIO to FTP
  • Ensures availability of all outputs before completing the task