What causes "job %q is in nonexistent node pool %q" in hashicorp/nomad?

Create the node pool first: nomad node pool apply (or the NodePools API), then submit the job.

job %q is in nonexistent node pool %q

Error message

job %q is in nonexistent node pool %q

What it means

The job endpoint validation hook verifies that every job references an existing node pool. It looks the pool up in state via NodePoolByName; if the lookup succeeds but returns nil, the job's cluster/nodes pool does not exist and registration is rejected.

Source

Thrown at nomad/job_endpoint_hook_node_pool.go:30

// jobNodePoolValidatingHook is an admission hook that ensures the job has valid
// node pool configuration.
type jobNodePoolValidatingHook struct {
	srv *Server
}

func (j jobNodePoolValidatingHook) Name() string {
	return "node-pool-validation"
}

func (j jobNodePoolValidatingHook) Validate(job *structs.Job) ([]error, error) {
	poolName := job.NodePool

	pool, err := j.srv.State().NodePoolByName(nil, poolName)
	if err != nil {
		return nil, err
	}
	if pool == nil {
		return nil, fmt.Errorf("job %q is in nonexistent node pool %q", job.ID, poolName)
	}

	return j.enterpriseValidation(job, pool)
}

View on GitHub (pinned to 482b49bf1a)

Solutions

Create the node pool first: nomad node pool apply <name> (or the NodePools API), then submit the job.
Fix the node_pool value in the job to an existing pool (nomad node pool list).
If the job should run in the default pool, remove the node_pool stanza so it defaults to "default".

Example fix

// before
job "web" {
  node_pool = "prod-eu" # does not exist
}
// after
# create it first: nomad node pool apply prod-eu
job "web" {
  node_pool = "prod-eu"
}

Defensive patterns

Strategy: validation

Validate before calling

// before submitting
pools, _, err := client.NodePools().List(nil)
if err != nil { return err }
want := job.TaskGroups[0].NodePool // or job-level node_pool
found := false
for _, p := range pools {
  if p.Name == want { found = true }
}
if !found { return fmt.Errorf("node pool %q not found; create it with `nomad node pool apply %s`", want, want) }

Type guard

func nodePoolExists(pools []*api.NodePool, name string) bool {
	for _, p := range pools {
		if p.Name == name {
			return true
		}
	}
	return false
}

Try / catch

// golang
_, _, err := client.Jobs().Register(job, nil)
if err != nil && strings.Contains(err.Error(), "nonexistent node pool") {
	// create the pool then retry once
}
// nomad CLI: check `nomad node pool list` before `nomad job run`

Prevention

Run nomad node pool list and cross-check job node_pool values in CI.
Provision node pools (Terraform/API) before deploying jobs.
Default to the "default" pool unless the pool is guaranteed to exist.
Keep pool names in shared config variables to avoid typos across clusters.

When it happens

Trigger: Calling the Job.Register/Validate API (or nomad job run) with a job whose node pool (job { node_pool = "..." }) names a pool that has not been created via the NodePools API.

Common situations: Typo in node_pool name; applying a job before the node pool Terraform/API provisioning; copying jobs between clusters where the target cluster lacks the pool; enterprise multi-cluster deployments where one cluster has the pool and another doesn't.

Related errors

AI-assisted analysis of hashicorp/nomad@482b49bf1a (2026-09-04). Data as JSON: /api/errors/b5bed827607fe7d4. Report an issue: GitHub.