NodesMonitoring

NodesMonitoring is a node monitoring system with Grafana dashboards and Prometheus metrics collection. It allows you to track node latency (ping) and availability, as well as export metrics to Prometheus.

🚀 Quick Start

git clone https://gitlab.bubki.zip/d3156/NodesMonitoring/
cd NodesMonitoring
sh download_plugins.sh
# Optional: edit example configs in ./configs/*
docker compose up -d --build

After starting:

• Grafana is available at: http://127.0.0.1:13000/dashboards
• Default credentials:
  • Login: admin
  • Password: admin

Nginx Configuration for Grafana

1. Update docker-compose.yml:

Replace the URLs with your domain:

environment:
  GF_SERVER_ROOT_URL: http://localhost:13000/
  GF_SERVER_HTTP_EXTERNAL_URL: http://localhost:13000/

2. Configure Nginx

Add the following location block to your server section in nginx.conf to properly reverse-proxy Grafana under /grafana/:

    location /grafana/ {
        proxy_pass http://127.0.0.1:13000/;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_set_header X-Forwarded-Proto $scheme;
        proxy_redirect http://127.0.0.1:13000/ /grafana/;
   }

Notes: proxy_pass points to your Grafana container’s internal port. proxy_redirect ensures Grafana correctly rewrites URLs when accessed through /grafana/. X-Forwarded-* headers preserve client information and protocol.

📊 Dashboards

• Ping Node — displays node latency and availability metrics.

Example:

📝 Configurations

1. configs/MetricsModel.json

• statisticInterval — statistics collection interval (seconds)
• stopThreadTimeout — thread stop timeout (ms)

2. configs/PingNode.json

• ping_interval_sec — ping interval
• timeout_ms — ICMP response timeout
• icmp_payload — ICMP packet content
• nodes — list of nodes to monitor

3. configs/PrometheusExporter.json

• mode — operation mode: pull (Prometheus scrapes) or push (Push Gateway)
• pull_port — port for pull metrics
• push_gateway_url — Prometheus Push Gateway address
• job — Prometheus job name

Dataflow Diagram

The following diagram illustrates the flow of data and metrics inside the PluginCore architecture:

Description of the dataflow:

• ModelStorage — global storage of models/interfaces for plugins.
• PingNodeModel registers itself in ModelStorage.
• PingNodePlugin interacts with PingNodeModel to collect and process node data, and sends metrics to MetricsModel.
• MetricsModel also stores its data in ModelStorage.
• PrometheusExporterPlugin reads metrics from MetricsModel and exports them to Prometheus.
• PluginLoader orchestrates the plugins and ensures metrics are exported to Prometheus.
• Prometheus serves metrics to Grafana for visualization.

This diagram clarifies how data flows from node monitoring plugins through the models and exporters to the visualization layer.