nomad alloc status: Inspect a Nomad Allocation
nomad alloc status reports an allocations task states, recent events, and (with -stats) live resource usage.
When a task keeps restarting, alloc status shows the per-task event timeline that names the cause, like an OOM kill or a failed restart.
What it does
nomad alloc status prints the allocations desired and client status, each tasks state with a Recent Events table (Received, Task Setup, Started, Terminated, Restarting), and the node it runs on. -stats adds live CPU and memory usage for running tasks.
Common usage
nomad alloc status 8d5d
nomad alloc status -stats 8d5d
nomad alloc status -verbose 8d5d
nomad alloc status -json 8d5dOptions
| Flag | What it does |
|---|---|
| -stats | Show live CPU and memory usage for running tasks |
| -verbose | Full IDs and the complete event history |
| -json | Emit the raw allocation object as JSON |
| -short | Condensed output without the event table |
In CI
When a deploy reports an unhealthy alloc, alloc status -verbose is the fastest path to the reason. The Recent Events table is where exit codes, OOM kills, and restart loops are spelled out.
Common errors in CI
In the event table, "Terminated ... OOM Killed: true" means the task exceeded its memory limit; raise the memory in the resources stanza. "Failed: ... Failed to pull image" points at a registry or auth problem. "No allocation(s) with prefix or id ... found" means a wrong alloc ID; list them via nomad job status.