Designing and shepherding cross-cutting features around compute resource isolation and utilization.
- Tuesdays at 18:00 UTC (weekly (On demand)). Convert to your timezone.
Meeting notes and Agenda can be found here.
- Vishnu Kannan, Google
- Derek Carr, Red Hat
A working group that is a cross-SIG effort currently sponsored by sig-node and sig-scheduling with a focus on improving Kubernetes Compute Resource Management.
Features tackled in this workgroup typically span more than one of the following SIGs:
- SIG-Node
- SIG-Scheduling
- SIG-Autoscaling
- SIG-Instrumentation
- SIG-Storage
- SIG-Networking
Many advanced resource management topics span these functional areas, and this working group is intended to support collaboration. Prior to pursuing major changes in compute resource management, it is encouraged that community members share their ideas in this working group to reach broad consensus across the SIGs. Once consensus is reached, the enhancements are then managed by the appropriate SIGs based on impacted areas of code.
It is an open and active working group, and we always welcome new additions!
Topics include, but are not limited to:
- Support for Performance sensitive workloads (exclusive cores, cpu pinning strategies, numa)
- Integrating new hardware devices (gpus, fpga, infiniband, etc.)
- Improving resource isolation (local storage, hugepages, caches, etc.)
- Improving Quality of Service (Performance SLOs)
- Performance benchmarking
- APIs and extensions related to the features mentioned above
- ...