SDN Throwdown overview: http://www.sdnthrowdown.com/
name | linkedin profile |
---|---|
Jialu Sun | [https://www.linkedin.com/in/jialu-sun-cse/] |
Bo Cheng | [https://www.linkedin.com/in/bo-cheng/] |
Litong Shen | [https://www.linkedin.com/in/litong-amanda-shen-b378727a/] |
Presentation ppt is located in docs folder.
Problem Statement
- You are given access to a shared volatile network where conditions change without warning
- You are given management tool to interact with the network
- Your mission is to generate traffic, transport it across this network and receive it at the other side
Problem Statement: Step 1
- Visualize the network
- Use the tools to describe what is going on and describe it Topology, health, utilization, etc.
Problem Statement: Step 2
- Get some traffic across the network
- Demonstrate that your traffic is crossing the network
- This traffic can be anything: be creative
Problem Statement: Step 3
- Manage the network
- Use the tools to alter the network in near-real-time to optimize your path
- React to changes and utilization
- You must identify the characteristics that you are optimizing for
- A network managed by NorthStar.
- Shared core network of 8 P routers
- Each group has its own PE routers and NorthStar Controller.
- LSPs use Segment Routing.
- Network telemetry collected by the NorthStar Controller.
- NorthStar notifies failures in the network.
- A pair of VM's dedicated to each team; on the east and west sides of the network.
- Traffic generators that vary in order to cause background north-south interference.
- Links in the network will occasionally fail.
- Deployed listener cluster on K8s to avoid single point failure.
- Leveraging AWS to auto-scale up during the traffic.