How to Configure Networking for Multi-Node GPU Worker Nodes
Configure multi-node GPU worker nodes, including Multus, RDMA resources, and NCCL or RCCL.
Manage DigitalOcean Kubernetes Clusters
Generated on 13 Nov 2026
DigitalOcean Kubernetes (DOKS) is a Kubernetes service with a fully managed control plane, high availability, and autoscaling. DOKS integrates with standard Kubernetes toolchains and DigitalOcean’s load balancers, volumes, CPU and GPU Droplets, API, and CLI.
Configure Rail-Optimized GPU Fabric
Verify the RDMA network interfaces on multi-node NVIDIA B300 GPU worker nodes in MKC1, which connect to a rail-optimized GPU fabric, and attach them to Pods with a CNI plugin that preserves their IP configuration.
How to Use the Routing Agent in Kubernetes Clusters
Manage IP routes on Kubernetes worker nodes using the routing agent
How to Use the Node Readiness Controller in DOKS
DOKS uses a node readiness controller to gate pod scheduling until all required node components, including GPU drivers, are healthy.
How to Customize CoreDNS for Kubernetes Clusters
Customize CoreDNS settings to change system configuration and provide specific options for domains.
Enable Peer-to-Peer OCI Registry
Enable the peer-to-peer OCI registry plugin to reduce container image pull traffic across a DigitalOcean Kubernetes cluster.
How to Upgrade DOKS Clusters to Newer Versions
Upgrade clusters to newer patch versions or newer minor versions using the control panel or command-line interface.
Configure Node Remediation
Automatically replace unhealthy DOKS worker nodes based on Kubernetes node conditions and rules you define with the Data Plane Operator.
How to Configure Single Sign-On for Kubernetes Clusters
Configure single sign-on (SSO) so team members can authenticate to your DigitalOcean Kubernetes clusters using your identity provider and OpenID Connect (OIDC).
How to Set RoleBindings for Custom Roles
Fine-tune access to Kubernetes clusters using custom team roles.