Skip to content

Conversation

tellet-q
Copy link
Contributor

@tellet-q tellet-q commented Apr 28, 2025

Accumulate CPU usage stats for default (!) benchmark strategy. Currently there are 2 sources:

  1. telemetry reports from qdrant
  2. utime obtained from server machine (calculated as difference between utime before benchmark and utime after the benchmark, divided by 100 to get number of seconds).

https://docs.rs/procfs/latest/procfs/process/struct.Stat.html#structfield.utime

Example run:
https://github.com/qdrant/vector-db-benchmark/actions/runs/14702755957

@tellet-q tellet-q requested review from agourlay and KShivendu April 28, 2025 08:35
@agourlay
Copy link
Member

Where can I see the result of the run regarding CPU stats?

@tellet-q
Copy link
Contributor Author

@agourlay

I've added 2 new panels into existing CI Benchmarks dashboard

image

Right now there are just a couple of datapoints.

@tellet-q tellet-q merged commit ce8fe71 into master Apr 29, 2025
11 checks passed
@tellet-q tellet-q deleted the cpu-monitoring branch April 29, 2025 12:46
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants