PostgreSQL Automatic Failover (PAF)
PostgreSQL Automatic Failover (PAF) is a robust OCF resource agent designed to enhance PostgreSQL clustering by leveraging Pacemaker and Corosync for high availability. It provides seamless failover capabilities, allowing for automatic election and promotion of standby nodes in case of master failures, while maintaining clear administrative boundaries between PostgreSQL and Pacemaker configurations.
Features
Supports PostgreSQL versions 9.3 and higher, ensuring compatibility with modern deployments.
Automatically detects and responds to master node failures, promoting standby nodes to maintain availability.
Provides detailed status reporting of PostgreSQL instances (master, slave, stopped, etc.) to Pacemaker.
Allows fine-grained control over PostgreSQL configurations without automatic adjustments, empowering administrators.
Extensive testing on CentOS 6 and 7 ensures reliability across various production environments.