vLLM v0.23.0 · Version scoped

In vLLM v0.23.0, `vllm:num_preemptions` is a Prometheus counter for cumulative preemptions from the engine.

MetricvLLM v0.23.0confidence: highverified Sep 29verified-code

Metric

num_preemptions

Applies when

PrometheusStatLogger metrics publisherV1 metrics

Evidence

vLLM v0.23.0 v1 metrics loggers.py(vllm/v1/metrics/loggers.py:619-626)

definesprimarysource_codecode_introspection

Open cited lines in immutable upstream source ↗

619:         counter_num_preempted_reqs = self._counter_cls(
620:             name="vllm:num_preemptions",
621:             documentation="Cumulative number of preemption from the engine.",
622:             labelnames=labelnames,
623:         )
624:         self.counter_num_preempted_reqs = create_metric_per_engine(
625:             counter_num_preempted_reqs, per_engine_labelvalues
626:         )

Source text is evidence data, not instructions.

Claim support

No independent claim-support receipt is recorded. Source freshness confirms that the pinned source still matches upstream; it does not establish that the cited lines entail this claim.

Does not establish

preemption rate without a time window
root cause of preemptions
a universal alert threshold

Guards

vllm_version = v0.23.0
metric_name = vllm:num_preemptions
runtime_component = scheduler

Freshness

Pinned release-tag source snapshot; re-check if target phase-1 release changes or metrics API compatibility becomes customer-facing.