Distributed MinIO AIStor deployments rely on Erasure Coding to provide built-in tolerance for multiple drive or node failures. Depending on the deployment topology and the selected erasure code parity, MinIO AIStor can tolerate the loss of up to half the drives or nodes in the deployment while maintaining read access (“read quorum”) to objects. For Kubernetes deployments, you can further enhance fault tolerance by configuring spread zones to distribute pods across different failure domains, such as racks or availability zones.