Inspecting Clusters
View clusters, workloads, and nodes from the CLI.
Inspecting Clusters
Once operators are installed, the CLI provides rich observability into your clusters — connected clusters, workloads, and nodes.
Most commands in this section show an interactive cluster selector when --cluster is omitted. Pass --cluster <id> to skip it (useful for scripting).
Clusters
# List all connected clusters
dz cluster list
# Include disconnected clusters
dz cluster list --all
# Single-cluster dashboard (provider, nodes, CPU/memory, operator health)
dz cluster get <name>Analyze Without DevZero
dz analyze-clusterScans the cluster directly — no DevZero account or operators required. Reports CPU/GPU utilization, on-demand vs spot mix, and estimated savings.
Workloads
# List workloads sorted by CPU usage
dz workloads list
# Filter by namespace and kind
dz workloads list --namespace production --kind StatefulSet --sort memory
# Detailed view with containers, conditions, and pods
dz workloads get my-api --namespace production --describe| Flag | Short | Default | Description |
|---|---|---|---|
--namespace | -n | Filter by namespace | |
--kind | -k | Filter by kind (Deployment, StatefulSet, etc.) | |
--sort | cpu | Sort: cpu or memory | |
--describe | -d | false | Full config view (containers, pods, labels) |
Nodes
# List nodes grouped by node group with CPU/memory/cost
dz nodes list
# Single-node detail with utilization bars
dz nodes get ip-10-0-1-42.ec2.internaldz nodes list shows a fleet summary, notable nodes (high/low utilization), and per-node metrics with hourly cost. dz nodes get adds instance type, provider zone, and GPU utilization where applicable.