Month: March 2025

This post discusses a perplexing issue with node evictions in an Oracle RAC setup, where node crashes occurred due to a loss of connection on a private interconnect despite redundancy. It highlights troubleshooting steps and findings, particularly focusing on the cluster_interconnects parameter. Ultimately, the solution involved resetting this parameter to default values, enabling the use of Highly Available IPs for improved reliability and communication in the cluster.