Skip to content

nodus spot up

View Markdown

Request one Spot node; –wait prints its SSH command when ready

nodus spot up GPU [flags]
nodus spot up h200x8 --max-price 12 --wait
nodus spot up h100x1 --ssh-key ~/.ssh/id_ed25519.pub --wait
-h, --help help for up
--idempotency-key string Stable key to retry the same logical request across invocations
--local-wait duration Local capacity wait before fresh storage elsewhere (0s–15m, whole seconds) (default 5m0s)
--max-attempts int Maximum accepted startup attempts, including the first (1–10; --retry-startup) (default 3)
--max-price float Maximum USD per complete node-hour
--placement string Workspace placement: local_first (default), local_only or cheapest
--poll duration Time between capacity/readiness checks (default 30s)
--regions strings Allowed Nodus locations or region classes, as shown by spot prices (comma-separated)
--resume Resume --state with its original request, attempt limit and deadline
--retry-startup Retry confirmed pre-SSH capacity failures after cleanup; requires --wait and --state
--ssh-key stringArray SSH public key file (repeatable; default: saved account keys)
--state string Private acquisition state file; reuse the same command/path to resume (--retry-startup)
--timeout duration Maximum request/wait duration; timeout does not release the node (default 15m0s)
--wait Wait for capacity and SSH readiness
--workspace string Canonical Spot Workspace ID whose retained storage mounts at /workspace
--workspace-slot string Independent block storage slot (default: default); shared storage uses default
--context string Context from the config file to use
--org string Organization (selects the context for that org)
-o, --output string Output format: json or yaml
-p, --project string Project to work in
-v, --verbose Log each API request (never credentials)
  • nodus spot - Request and manage dedicated interruptible nodes (pilot)