Reports the power and enrolment state of a node set. Read-only: stat never
changes anything, so it is always safe to run.
omf stat -t <topology> [options]
| Option | Meaning |
|---|---|
-t, --topology TOPO |
the node set — see topology formats |
-s, --summary |
counts instead of per-node lines |
-d, --domain DOMAIN |
override the auto-detected sandbox |
-v, --verbose / --debug |
more output |
Per-node detail:
$ omf stat -t node1-4,node1-5
node1-4.sb4.cosmos-lab.org POWERON
node1-5.sb4.cosmos-lab.org POWERON
Counts across a whole sandbox — the fastest way to see what you are holding:
$ omf stat -t all --summary
Power ON: 9
Power OFF: 20
Nodes Not Available: 2
Admin Down: 2
Only the nodes from the last imaging run:
omf stat -t system:topo:imaged
omf stat -t system:topo:failed
| State | Meaning |
|---|---|
POWERON |
the chassis manager reports the node is powered |
POWEROFF |
powered down |
Admin Down |
administratively disabled — not yours to use |
Nodes Not Available |
not in your reservation, or not present |
Unknown (UNREACHABLE) |
the CM itself is not answering |
POWERON is not "the node is usable"This is the most common misreading. omf stat asks the chassis manager
whether power is applied. It does not ask the node anything. A node can report
POWERON while:
To know what a node is actually running, ask the node:
ssh root@node1-4 'cat /.omf-image-name; uptime -p'
Unknown (UNREACHABLE) means the CM is dead, and that is dangerousWhen the chassis manager cannot report state, power commands become
unpredictable rather than merely unavailable: omf tell offh can still power
the node off while replying ERROR, and on then also fails — leaving the
node stranded until someone attends to it physically.
If a node reports Unknown (UNREACHABLE), do not issue power commands to it.
Check whether it is on the known-bad-CM list first and ask.
Author(s): COSMOS team · Last verified: 2026-09-11 — options from omf stat --help; outputs are real, taken on sb4 and grid consoles · Tags: omf, user-guide, reference