In vLLM v0.23.0, `vllm:request_queue_time_seconds` use quantiles with running/KV/token-progress metrics to distinguish admission delay from decode delay; no universal threshold is established.

MetricvLLM v0.23.0confidence: highverified 11 hr agoverified-code

PrometheusStatLoggerV1 metrics

vLLM v0.23.0 v1 metrics stats.py(vllm/v1/metrics/loggers.py:406-455)

supportsprimarysource_codecode_introspection

Pinned release-tag semantics; do not carry to another version without continuity evidence.

In vLLM v0.23.0, vllm:request_queue_time_seconds use quantiles with running/KV/token-progress metrics to distinguish admission delay from decode delay; no universal threshold is established.