Purpose-built runtime, Kubernetes-only focus
Day 1
Architecture, Configuration & Operations
- CRI-O vs containerd: Positioning, scope, trade-offs — when CRI-O is the right choice
- CRI-O architecture: The runtime stack from kubelet to OCI runtimes — how CRI-O fits in
- OCI runtime configuration: runc as default, crun for improved performance — configuring the low-level runtime
- Image handling: Pulling, storing, and managing images with containers/image — the image subsystem
- CRI-O configuration: /etc/crio/crio.conf — runtime, storage, network settings explained
- Pinning OCI images: Reproducible node configuration with pinned image digests
- Registry configuration: Insecure registries, registry mirrors, auth files — connecting to private registries
- Network: CRI-O with CNI plugins — Flannel, Calico, OVN-Kubernetes integration
- crictl: Inspecting pods, containers, and images in a running cluster — the primary debug tool
- Log management: CRI-O logging with journald and file backends
- OpenShift: CRI-O as the default runtime — configuration through MachineConfig operators
- Upgrading CRI-O safely alongside Kubernetes: Version compatibility and upgrade sequencing
What your team walks away with
A solid understanding of CRI-O's focused runtime model, how to configure it for production nodes, and how to debug container lifecycle issues in OpenShift and vanilla Kubernetes.
- Configure CRI-O for production nodes: registries, runtime, storage
- Debug container lifecycle issues using crictl — pods, containers, images
- Tune CRI-O for performance with crun and appropriate storage drivers
- Manage CRI-O in OpenShift through MachineConfig operators
Book the CRI-O training
Recommended for OpenShift administrators and Kubernetes node operators who need to understand the runtime layer beneath their workloads.
Get in touch