Brigade Goals and Scope
The overarching goal is to optimize the CORD platform and benchmark performance for several critical workloads. Specifically, the plan is to:
- Integrate Intel's DPDK/OvS optimizations into the CORD platform
- Initial Focus: VMs
- Follow-on Work: Containers
- Micro-Benchmarks / Tuning
- Cores dedicated to OvS vs OvS/DPDK
- Benchmark M-CORD performance (specifically EPC)
- Scale number of subscribers (at average throughput each)
- Maximum end-to-end throughput for a single subscriber
- Benchmark R-CORD performance (specifically vSG)
- Scale number of subscribers (at average throughput each)
- Maximum end-to-end throughput for a single subscriber
- Open Ended: Other optimizations that impact the M/R-CORD performance
How to get involved
Please join the brigade in their #performance Slack channel to discuss what is happening and how you can participate. If you need access to the CORD slack channel, you can request an invitation.
Brigade Leads
- Sai Edupuganti, Intel <saikrishna.edupuganti@intel.com>
- Sisir Chowdhury, Radisys <Sisir.Chowdhury@radisys.com>
- Chetan Gaonker, Ciena <chgaonke@partner.ciena.com>
Mentors
- Andy Bavier, ON.Lab <andy@opennetworking.org>
Product Owner
- Tom Tofigh (AT&T)
Brigade Members
- Anil Kumar Sanka <anilkum@partner.ciena.com>
- Naresh Sugumar <naresh.sugumar@spirent.com>
Events / Meetings / Forums
- meet.google.com/ojv-otmf-uqy Fridays at 9am Pacific time
Resources
