This blog post delves into deploying Redpanda for high availability (HA) within a single availability zone (AZ), emphasizing the importance of eliminating single points of failure by spreading replicas across multiple failure domains or racks. It examines the pros and cons of single AZ deployments compared to multi-AZ deployments, noting that while single AZ setups avoid cross-AZ network charges, they rely on robust local infrastructure to prevent failures. The post also explains how Redpanda's rack awareness and the Raft consensus algorithm facilitate efficient HA deployments without significant cost or complexity increases. It outlines infrastructure considerations and describes how to automate deployments using Terraform and Ansible across AWS, Azure, and Google Cloud, with each cloud provider offering specific tools for spreading hosts across failure domains. Additionally, it provides a practical example of setting up a highly available Redpanda cluster, highlighting the configuration of rack awareness and the use of automation scripts to ensure seamless deployment. The series will continue by addressing HA clusters across multiple availability zones and regions.