Skip to content

Conversation

@tommy-u
Copy link
Contributor

@tommy-u tommy-u commented Aug 19, 2025

Add cgroup_cli.sh, a simple CLI tool wrapping systemd-run for starting, stopping, and monitoring synthetic workloads. This tool is useful for building microbenchmark experiments that stress particular scheduler functionality. See more in README.md.

2 example experiments are included. One illustrates launching 10 cpu hog threads across 4 cgroups each with 10 CPUs. Another demonstrates launching 2 cpu hogs across 2 CPUs which are chosen to be on different LLC slices on a given architecture.

Add cgroup_cli.sh, a simple CLI tool wrapping systemd-run for
starting, stopping, and monitoring synthetic workloads. This tool
is useful for building microbenchmark experiments that stress
particular scheduler functionality. See more in README.md.

2 example experiments are included. One illustrates launching 10
cpu hog threads across 4 cgroups each with 10 CPUs. Another
demonstrates launching 2 cpu hogs across 2 CPUs which are chosen
to be on different LLC slices on a given architecture.
@tommy-u tommy-u requested review from dschatzberg and kkdwvd August 19, 2025 23:39
@kkdwvd
Copy link
Contributor

kkdwvd commented Aug 20, 2025

Looks good in general, but can we have a single script that takes different arguments for different combinations?

@tommy-u
Copy link
Contributor Author

tommy-u commented Sep 3, 2025

Sorry @kkdwivedi, just back from PTO. Combining synthetic_workloads/expts/* into one script? Sure thing!

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