Skip to content

Add request cpu time prometheus metrics

Qingyu Zhao requested to merge add-request-cputime-prometheus-metrics into master

What does this MR do?

For issue #31890 (closed)

This is to have prometheus histogram of different endpoints with their distribution of CPU-time.

Having request execution time (time spend in thread) will allow us to see which requests cause severe performance impact (a lot of time spend in Ruby thread) and cause contention on Puma multi-threaded model.

This MR should be merged after !18498 (merged) to reuse the new change in Gitlab::Metrics::System.thread_cpu_time(refer to !18498 (comment 230588579)).

Edited by 🤖 GitLab Bot 🤖

Merge request reports