Distributed PostgreSQL Cluster

This repository implements a Distributed PostgreSQL Database architecture designed for High Availability (HA) and resilience, fully orchestrated within Docker.

The core objective of this project is to validate how a distributed system handles critical failures. Instead of theoretical setups, I focused on Resilience Testing by:

1. Manual Failure Simulation: Forcibly stopping Docker containers (Master/Slave nodes) to trigger real-time Failover and Failback events via repmgr.
2. Traffic Stability Analysis: Running concurrent stress tests during these outages to verify that Pgpool-II successfully reroutes queries and maintains service availability with minimal downtime.
3. Observability: Utilizing a full monitoring stack (Prometheus, Grafana, Postgres Exporter) to visually track replication lag and cluster health during the chaos.

Setup

1. Run the containers

docker compse up -d

2. Open your grafana at http://localhost:3000 and login with password in docker-compose.yml
3. From dashboard, open menu -> Connections -> Data Sources
4. Set Connection to http://prometheus:9000 and click Save & Test
5. Back to dashboard click new button -> import.
6. at Grafana dashboard URL or ID set value 9628 (postgreSQL).
7. Click load and the dashboard ready to use