prometheus / prometheus/node_exporter
optimize cpufreq collector for multithreaded physical cores
Nobody has claimed this yet.
- Dominant language
- Go
- Stars
- 13.8k
- Forks
- 2.7k
- Avg merge
- 1d 23h
- Merged PRs (30d)
- 8
Description
(enhancement)
A proposal for performance improvement at the node_exporter cpufreq collector. The frequencies for each "CPU" are read always, despite of having multiple threads running on the same physical core (aka "hyperthreading").
If two CPUs share the same physical core, they share the frequency information. It would be possible to just copy the info instead of reading sysfs for both, based on the topology info and the thread siblings list. The only fields that come to my mind which are different, based on what I've seen, would be SystemCPUCpufreqStats Name and RelatedCpus.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start at the node_exporter cpufreq collector and inspect how it reads per-CPU sysfs data. Compare that flow with the topology information and thread-siblings list, while preserving the distinct SystemCPUCpufreqStats Name and RelatedCpus fields. Done means avoiding duplicate frequency reads for CPUs sharing a physical core without changing the reported metrics.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go, linux
- Domain
- operating-systems, performance
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 38/100