does anyone knows this metris(grpc_responses_duration_seconds_bucket) means? #5869
Replies: 1 comment 1 reply
|
Those are components of a Prometheus classic histogram for gRPC response duration, measured in seconds:
In your Upsert sample, 6 observations were ≤ 0.00005 s (50 µs), and 34 were ≤ 0.0005 s (500 µs). Thus 28 were in Your For a recent mean by endpoint across instances: For an estimated p95 (seconds), using all exposed buckets: Add your service/job selectors if Prometheus scrapes multiple Qdrant deployments. These are server instrumentation timings, not automatically the client's end-to-end latency. References: Qdrant API response metrics, Prometheus histogram interpretation. |
Uh oh!
There was an error while loading. Please reload this page.
hello, everyone,
do you know "grpc_responses_duration_seconds_bucket" means? I found this from /metics
grpc_responses_duration_seconds_sum{endpoint="/qdrant.Points/Get"} 2.654579
grpc_responses_duration_seconds_count{endpoint="/qdrant.Points/Get"} 11633
grpc_responses_duration_seconds_bucket{endpoint="/qdrant.Points/Upsert",le="0.00001"} 0
grpc_responses_duration_seconds_bucket{endpoint="/qdrant.Points/Upsert",le="0.00005"} 6
grpc_responses_duration_seconds_bucket{endpoint="/qdrant.Points/Upsert",le="0.0001"} 6
grpc_responses_duration_seconds_bucket{endpoint="/qdrant.Points/Upsert",le="0.0005"} 34
grpc_responses_duration_seconds_bucket{endpoint="/qdrant.Points/Upsert",le="0.001"} 45
All reactions