Skip to Content
View all events

Designing Extensible Data Center Fabrics

Published by Pearson

Intermediate content levelIntermediate

Understanding the theory and design of large-scale spine-and-leaf data center fabrics and their control planes.

  • Learn the history and background of spine-and-leaf fabrics, points informing their optimal design and use
  • Get detailed information on scaling, lifecycle, device selection, and other aspects of data center fabric design
  • Explore data center fabric control planes and their tradeoffs, with the implications of control plane choices on security, automation, and other operational aspects

Data centers are the foundation of the cloud, whether private, public, on the edge, or in the center of the network. This training will focus on topologies and control planes, including scale, performance, and centralization. This training is important for network designers and operators who want to understand the elements of data center design that apply across all hardware and software types.

This class consists of two three-hour sessions. The first session will focus on the physical topology, including a short history of spine-and-leaf fabrics, the characteristics of fabrics (versus the broader characteristics of a network), and laying out a spine-and-leaf network to support fabric lifecycle and scaling the network out. The first session will also consider the positive and negative aspects of using single- and multi-forwarding engine (FE) devices to build a fabric, and various aspects of fabric resilience.

The second session will begin with transport considerations and quality of experience. The session will then consider underlay control planes, including BGP and IS-IS, and the positive and negative aspects of each. Routing to the host and the interaction between the control plane and automation will be considered in this session, as well. EVPN as an overlay control plane will be considered next, and finally the relationship between security and control plane design will be examined.

What you’ll learn and how you can apply it

  • The advantages and disadvantages of using spine-and-leaf fabrics for data center networks
  • The data center fabric lifecycle
  • The advantages and disadvantages of different control plane options
  • Telemetry challenges and potential solutions

And you'll be able to:

  • Lay out a spine-and-leaf fabric for optimal lifecycle and scaling
  • Design a data center fabric that works with security and automation systems
  • Determine where and how to layer and deploy routing for optimal performance

This live event is for you because...

  • You want to learn data center spine and leaf topologies, including variants, scaling, and theory
  • You want to learn how to optimally use BGP as a data center control plane
  • You want to learn about link state protocols for data center fabrics
  • You want to learn about performance metrics for data center fabrics
  • You want to learn about telemetry challenges for data center fabrics

Prerequisites

  • Understand control planes, particularly BGP and IS-IS
  • Understand IPv4, IPv6, and MPLS
  • Understand basic automation and security concepts

Recommended Preparation

Recommended Follow-up

Read: The Art of Network Architecture by Russ White and Denise Donohue

Schedule

The time frames are only estimates and may vary according to how the class is progressing.

Session 1 (3 hours)

Segment 1: Traditional hierarchical design (50 minutes)

  • The traditional DC network design
  • Traditional design challenges
  • Moving to a fabric
  • How the fabric solves the challenges to traditional designs

Q&A (5 minutes)

Break(5 minutes)

Segment 2: The spine-and-leaf fabric (50 minutes)

  • Crossbar, Strowger, Clos, folded Clos
  • The butterfly fabric (five-stage)
  • Laying out the spine-and-leaf fabric

Q&A (5 minutes)

Break (5 minutes)

Segment 3: Fabric lifecycle and resilience (50 minutes)

  • The importance of the pod
  • Managing changes
  • Resilience and the law of large numbers
  • Testing and learning resilience

Q&A (10 minutes)

Session 2 (3 hours)

Segment 4: Transport and Underlay Considerations (50 minutes)

  • Optics and breakouts
  • Quality of experience
  • BGP as an underlay
  • IS-IS as an underlay

Q&A (5 minutes)

Break (5 minutes)

Segment 5: The eVPN overlay (50 minutes)

  • Layered control planes and managing state
  • The problems to be solved
  • eVPN as a solution

Q&A (5 minutes)

Break (5 minutes)

Segment 6: Traffic Steering and Security (50 minutes)

  • Traffic steering use cases
  • Traffic steering in spine-and-leaf fabrics
  • Security and failure domain separation through layered control planes
  • Thoughts on the attack surface

Conclusion (10 minutes)

Your Instructor

Russ White

Russ White has experience in designing, deploying, breaking, and troubleshooting large scale networks, and is a strong communicator from the white board to the board room. He has co-authored more than forty software patents, participated in the development of several Internet standards, helped develop the CCDE and the CCAr, and worked in Internet governance with the Internet Society. Russ has a background covering a broad spectrum of topics, including radio frequency engineering and graphic design, and is an active student of philosophy and culture. Russ is a co-host of the Hedge podcast, serves on the Routing Area Directorate and the Internet Architecture Board at the IETF, co-chairs the BABEL working group, and serves on the Technical Services Council/as a maintainer on the open source FR Routing project. His most recent works are Computer Networking Problems and Solutions and Unintended Dystopia. Russ regularly teaches live webinars on Internet technology through Safari Books Online, as well. MSIT Capella University MACM Shepherds Theological Seminary PhD, Southeastern Baptist Theological Seminary CCIE #2635, CCDE 2007::1, CCAr

linkedinXlinksearch

Skill covered

Data Center