medmfc-monitoring

Purpose

The central monitoring platform receives node telemetry, stores canonical cycle/event state, performs independent MDS/S3 confirmation, and exposes the operator dashboard.

Repository tree

medmfc-monitoring/
├── .env.example
├── .gitignore
├── README.md
├── docker-compose.yml
│
├── backend/
│   ├── Dockerfile
│   ├── requirements.txt
│   └── app/
│       ├── __init__.py
│       ├── event_log.py
│       ├── main.py
│       ├── s3_monitor.py
│       ├── s3_validator.py
│       └── settings.py
│
├── frontend/
│   ├── Dockerfile
│   ├── index.html
│   └── nginx.conf
│
└── docs/
    ├── deployment.md
    ├── event-schema.md
    └── status-model.md

Runtime architecture

flowchart LR
    AG[HCMR Agent] -->|Authenticated telemetry| API[FastAPI Backend]
    API --> DB[(PostgreSQL)]
    API --> LOG[(JSONL / Human Logs)]

    S3[Public MDS / S3] --> WORKER[S3 Monitor]
    WORKER --> API

    DB --> API
    API --> WEB[Frontend / nginx]
    WEB --> OP[Operator]

Current responsibilities

receive authenticated agent events and state;
persist canonical event and cycle data in PostgreSQL;
maintain append-only operational logs;
expose current-cycle and event APIs;
perform background MDS/S3 validation;
promote confirmed cycles to MDS_CONFIRMED;
expose the dashboard and continuous event log;
provide the basis for TDT, alerting and duplicate-delivery detection.

Persistent host state

Runtime data lives outside the Git checkout:

/home/nocuser/medmfc-monitoring-data/
├── postgres/
├── logs/
│   ├── app/
│   ├── events/
│   └── human/
├── exports/
├── docs/
├── cache/
└── backup/

Deployment model

The repository is deployed as a Portainer Git stack from the main branch using docker-compose.yml.

Secrets remain in Portainer/environment configuration and are not committed to Git.