Machine Status Reference
The Status tab summarizes an environment's machine activity end to end, and the Machines list shows each machine in detail.
These are driven by best-effort machine status report and heartbeat data. During network interruptions or other systemic issues, the data may be out of date or incorrect.
Status summary
Cards at the top of the Status tab. Counts labelled "last 30 days" cover the machines still in the retained report history.
Headline counts
Active machines
Machines currently live.
The sub-line splits them into busy (running a job), ready (online but idle), and initializing (still starting up).
Stuck machines
Live machines that may need attention.
Stalled, or reporting an inconsistent state. Turns red above zero.
Failed inits (last 30 days)
Machines whose setup failed before the worker came online.
Stopped on a failed Provisioning or Initializing step. Turns red above zero.
Failed terminations (last 30 days)
Terminated machines that ended in an error.
Their final Running or Stopping step errored. Turns red above zero.
Median time per lifecycle state (last 30 days)
A bar showing where machine-time goes — the median time machines spend in each state, in order:
Requested → Provisioning → Initializing → Ready → Busy → Stopping.
Ready and Busy split the Running phase: Ready is time online waiting for work, Busy is time running jobs. A state appears only once there is data to measure it, and the total is the sum of the per-state medians.
Idle time % (last 30 days)
The share of operational time (Ready + Busy) that machines spent idle. Reads "No operational time recorded" until machines have run.
Scaling summary
The last scaling time and its result, as reported by the scaling infrastructure.
Recent events
The environment's scaling activity log, newest first.
Machines list
Activity timeline
Machines report events as they start, run jobs, encounter errors, and stop. Events appear in each machine's activity timeline.
Events that report setup, worker, or queue failures are marked as errors. Check the environment configuration or contact Valohai support if errors persist.
Lifecycle phases
Order: Requested → Provisioning → Initializing → Running → Stopping.
Requested
VM requested, not yet booting.
Provisioning
Host booting and setting up.
Initializing
Host ready, worker agent starting.
Running
Worker operational — idle or running a job.
Stopping
VM stopping or terminated.
Running is labelled Idle when the machine has no job, Running when it is working.
Machine states
One state per machine.
Active
Live; working, starting up, or activity unknown
Working or coming online.
Idle
Live; Running with no job
Operational, waiting for work.
Stuck
Live; no status report for 15+ min, or unresolved error in the current phase
Needs attention.
Historical
Terminated
Machine gone; report retained.
States are mutually exclusive.
A machine that hit an error but recovered is not Stuck.
See Also
Hardware Statistics — CPU, memory, and GPU utilization for executions
Configure Environments & Scaling — Machine types, autoscaling, and scaling behavior
Last updated
Was this helpful?