vLLM KV-consumer fetch gauge by stage
vllm · gauge
Fields
id
metric-exposures/w3m-vllm-vllm-num-requests-kv-fetch-by-stageMetric name
vllm:num_requests_kv_fetch_by_stage
Metric type
gauge
Unit
requests
Exposure
prometheus
Endpoint
:8000/metrics
Labels
model_nameenginestage
Enabled by default
yes
notes
vLLM main, PrometheusStatLogger in vllm/v1/metrics/loggers.py, read 2026-10-04. On by default (EngineArgs.disable_log_stats defaults False); served at :8000/metrics. Bucket family name in --custom-histogram-buckets: no bucket family. Only emitted when the process runs as a KV cache consumer, so its absence on a standalone server is expected.
Notes
vLLM main, PrometheusStatLogger in vllm/v1/metrics/loggers.py, read 2026-10-04. On by default (EngineArgs.disable_log_stats defaults False); served at :8000/metrics. Bucket family name in --custom-histogram-buckets: no bucket family. Only emitted when the process runs as a KV cache consumer, so its absence on a standalone server is expected.