Building CLOS Fabrics: The Complete Handbook: Design, Deploy, and Troubleshoot Spine-Leaf Networks with BGP, EVPN-VXLAN, and Automation. Real-World Labs Included - Tapa blanda

VEYSON, EIRA

 
9798269851471: Building CLOS Fabrics: The Complete Handbook: Design, Deploy, and Troubleshoot Spine-Leaf Networks with BGP, EVPN-VXLAN, and Automation. Real-World Labs Included

Sinopsis

Ship a spine leaf Clos fabric that scales, survives failure, and stays easy to operate.

Designing a spine leaf network is straightforward in theory, yet teams struggle with capacity math, stable underlays, and an EVPN VXLAN control plane that holds up under change. This book guides you from first principles to a working, scalable fabric that you can operate with confidence.

Across twelve focused chapters, you will plan port counts and oversubscription, deploy eBGP everywhere with IPv6 only transport, implement EVPN VXLAN with symmetric IRB and anycast gateways, and validate everything with automation and telemetry.

  • Clos and fat tree capacity planning, oversubscription math, pod sizing, ECMP width
  • Underlay eBGP design, per tier AS strategies, IPv6 only with RFC 8950 next hop
  • Security on day one, GTSM, TCP AO or MD5, reasonable CoPP baselines
  • EVPN route types 1 to 6, VNI planning, ARP or ND suppression, IMET and SMET
  • Anycast gateway and IRB models, symmetric versus asymmetric trade offs, type 5 routing
  • EVPN multihoming, ESI planning, DF election options, split horizon and local bias
  • BUM control choices, ingress replication limits, multicast underlay with PIM
  • DCI and multi site patterns, EVPN to EVPN or IP VPN interconnect, RFC 9014 gateways
  • Automation with NetBox and Ansible AVD, deterministic renders, golden templates and CI
  • Change safety, Batfish preflight invariants, canary rollout, rollback routines
  • Observability with OpenConfig gNMI, BMP based BGP monitoring tied to SLOs
  • Performance engineering, ECMP hashing verification, convergence timers, MTU and ECN
  • RoCE tuning for AI or storage, PFC, DCQCN, watchdog practices that prevent stalls
  • Physical layer choices, DAC and AOC, SR DR FR LR optics, PAM4 and FEC considerations
  • Real world labs, failure drills for link flap, node loss, process restart, DF flips

This is a code heavy guide. You get working Containerlab topologies, FRRouting and EOS snippets, NetBox and AVD data models and playbooks, Batfish queries, and traffic generator profiles that let you validate behavior end to end.

Get the practical reference you will keep open on deployment day and during incident review, grab your copy today.

"Sinopsis" puede pertenecer a otra edición de este libro.