Data Center Fabric Architectures

Overall rating: 4.68 Instructor: 4.85 Materials: 4.77 more …

This webinar describes the basics of Data Center fabric architectures offered by numerous vendors including Alcatel Lucent, Arista Networks, Avaya, Brocade, Cisco, Force10 Networks (now Dell), HP, Juniper, Cumulus, NEC and Plexxi.

Regular update sessions describe the progress individual vendors made since the initial webinar was released (check also the list of webinar videos and materials).

We stopped tracking smaller vendors and vendors focusing on reselling third-party hardware of network operating systems. Please check the current webinar contents for details.

Availability

This webinar is part of Data Center Infrastructure roadmap and accessible with standard subscription

Start now Access content

Contents

This webinar will help you select the best solution for your data center switching needs. It starts with:

  • Introduction and market overview
  • Data center fabric requirements
  • Technology overview, including merchant silicon, SoC architectures, speeds-and-feeds (including 400 GbE), cable breakout options, and Universal Forwarding Tables

Next, the webinar focuses on vendor-specific fabric architectures and classifies them based on control and data plane position within the fabric into the following major groups:

  • Independent management, control and data plane – Multi-chassis Link Aggregation (MLAG) and large-scale bridging solutions from ALU, Arista, Avaya, Brocade, Cisco and Force10.
  • Centralized management, independent control and data planes – Cisco’s UCS Manager, Brocade VCS Fabric, Juniper’s Virtual Chassis and QFabric, Plexxi
  • Centralized control plane, distributed data planes - Cisco’s VSS, HP’s IRF, Juniper’s XRE, NEC ProgrammableFlow.
  • Centralized control plane with offloads - Juniper Virtual Chassis Fabric, Big Cloud Fabric
  • Centralized data plane – Cisco Nexus Port Extenders, Juniper Fusion

Majority of the webinar content describes hardware characteristics and software features of major data center switching platforms, and explains how well each vendor platform/product fulfills the following criteria:

  • Optimal bandwidth utilization;
  • Redundant edge connectivity and Multi-chassis Link Aggregation (MLAG);
  • Layer 2 and Layer 3 multipathing;
  • Optimal layer-3 forwarding;
  • Multi-tenant support and multiple routing domains;
  • Virtualization awareness;
  • Storage protocol integration including FCoE support;
  • Data Center Bridging (DCB) support;
  • Large forwarding table sizes and deep packet buffers;
  • Programmability, simplified provisioning and management.

The foundation technologies described in this webinar include:

400 GbE

The exponential growth in data and IP services are pushing the high-speed optics industry for continuous innovations to address current and future technology and cost demands. This part of the webinar describes the available implementations of 400GbE (including a bit of 25 GbE, 50 GbE, and 100 GbE), key technology enablers, solutions and their impact.

Engineers working for Cloud, Enterprise and SP companies will get the in-depth information on using 400GbE in Data Center environments, and a bit of WAN applications.

Switching ASIC Deep Dive

Network engineers understand that ASICs are the magical heart of data center switches, but few of us understand how they work. Without violating any NDAs, the session with Pete Lumbis will shine some light on what ASICs are, how they operate and what their strengths and limitations are. Pete will also briefly discuss single chip and multi-chip (chassis) systems and how they differ as well as the role buffers and table sizes play in the land of ASICs.

Upcoming Update Sessions

You'll find the details of prior update sessions in the update sessions tab.

Update Sessions

The initial webinar describing fabric architectures from numerous vendors is augmented with annual update sessions. The update sessions document the progress and changes to product lines and architectures the vendors made in the last 6-18 months. We covered these topics in our update sessions:

2020

On April 23rd 2020, Pete Lumbis explained the mysteries of switching ASICs from a Network Engineer perspective.

In March 2020 Lukas Krattiger and Mark Nowell described the technology foundations and use cases for 400 GbE in Data Center and WAN.

2019

In October 2019 Pete Lumbis described new features in Cumulus Linux including:

  • New ASIC support including (Broadcom's Trident3, Tomahawk2 and Tomahawk3, Mellanox Spectrum II)
  • Cumulus Linux 4.0 Update and the move to Debian Buster
  • Simplifying your network operations with Auto BGP, Auto MLAG and smart data center defaults
  • Linux streaming telemetry with Cumulus NetQ
  • Sneak Peek: Model-driven configuration

2018

  • New webinar introduction
  • Market overview
  • Hardware overview covering switch-on-a-chip (SoC), fixed and modular switches
  • Merchant silicon versus vendor ASICs
  • 10 GE - 100GE speeds and breakout options
  • How to choose the best solution for your needs
  • Hardware overview, merchant silicon used, and EOL models of Arista, Cisco and Juniper data center switches

2017

  • New Arista switches and enhancements in EOS releases 4.17 and 4.18
  • New Cisco Nexus switches and new features in Nexus OS for Nexus 9000 switches
  • Brocade SLX series and new features in Network OS 7.1
  • New features in Cumulus Linux (presented by Dinesh Dutt)

2016

  • Redesigned Requirements section
  • Redesigned Fabric Architectures sections
  • New hardware and software features from Arista, Brocade and Cisco
  • New hardware and software features from Dell, HP and Juniper

May 2015

  • New 100GE switches from Arista and VXLAN support in Arista EOS;
  • VDX 6940 switches from Brocade and new Network OS features including Fabric Virtual Gateway;
  • Cisco Nexus 9000-series switches and new Nexus OS features including VM tracker and VXLAN support;
  • New switches from Dell Force10 and their VXLAN and OpenFlow support;
  • HP Data Center switches and OpenFlow support;
  • Overview of Juniper QFX10000 switch

May 2014

  • New Spline switches from Arista and new Arista EOS features;
  • A whole new look on Brocade VCS fabric, including L2 and L3 forwarding, logical chassis and virtual fabrics, STP and LACP integration, and VMware NSX VXLAN gateway;
  • New Cisco Nexus switches, an overview of Dynamic Fabric Automation, and description of anycast HSRP and routing protocol peering over vPC+
  • A completely overhauled Dell Force10 section;
  • An overview of data center protocols in various HP data center switches;
  • Rewritten description of Juniper's Virtual Chassis and enhancements introduced in Virtual Chassis Fabric.

May 2013

Dan Backman, solution architect @ Plexxi, described Plexxi's approach to scale-out data center fabrics and controller-based data center networking (more about Dan).

Other highlights of this update session include:

  • In-depth description of new Arista EOS features, including IPv6 support, EOS API, Rapid Indication of Link Loss (RAIL), and tap aggregation on Arista 7150;
  • New data center switches from Cisco including Nexus 6000;
  • Juniper EX 9200 and new Junos 12.3 features.

November 2012

  • Arista 7150 series switches and new EOS features in releases up to 4.10.3
  • Brocade VDX 8770 switch and new features in Network OS 3.0
  • Cisco Nexus 7000 additions, Nexus 3548, port extenders overview and new features in Nexus OS
  • Juniper QFabric (QFX3000-M) and new Junos 12.2 features

May 2012

The update session focused on the following products/architectures:

  • Cisco Nexus series
  • HP Intelligent Resilient Framework
  • Juniper Virtual Chassis

The session described solutions from nine major vendors: Alcatel Lucent, Arista, Avaya, Brocade (VDX series only), Cisco (Nexus series only), Dell Force 10, HP, Juniper and NEC.

In addition to the detailed product data for each individual vendor, four new criteria were added to the solution scorecards:

  • Optimum Layer 3 forwarding across the fabric;
  • Layer 3 path isolation (based on MPLS/VPN or VRF-like multiple routing tables);
  • IPv6 support;
  • Hardware tables sizes (MAC, IPv4, IPv6, IP multicast, ARP and ND).

Target Audience

This webinar is ideal for IT managers and networking engineers that have to look behind the vendor marketectures to understand the benefits and drawbacks of individual data center solutions.

If you are a network architect, designer or sales/support engineer working in data center environment, you simply have to attend this webinar.

Happy Campers

About the webinar

The added value is the comparison between different vendors, to do this yourself you need a lot of time which I haven't. You need also accurate info, which is not always available via Internet, but which Ivan is able to get via his relations. Not everyone is able to give insights over so many topics, only very experienced people can.

Guy Van De Wiel
The most valuable part of this webinar is keeping up with what all the vendors have been working on in the past 6 months.
Bryan
All you need to know about data center fabrics using major vendors of market. Wonderful summary. I liked the idea or rating system you put for each vendor and their products families.
Krunal Shah
Excellent! As usual...
Pierre-Louis Gingembre
Yes
John Lebrun
I've been watching these since the very beginning and they are invaluable both for initial learning and for subsequent reference. Fantastic interview preparation too ;) Bargain!
James Smith
Good information, good presentation.
Marc Enders
Great summary of network fabrics available on the market with bonus including explanation how new things works on those fabrics. Very useful if you are searching for information to do decision what to buy.
Matej Rehak
If you want straight forward technology updates, based on real features, you should always include Ivan's webinars.
Hagen Amen
Good look into the confusing space of how whitebox, disaggregation and SDN relate and what they really bring to the table. A deeper look into the tradeoffs that need to be made when you start down this path.
Christopher Young
I like what you bring for us
Hassan HBAR
Excellent update on Cumulus. Great take away was existence of FRRouting. Will be experimenting with it ASAP.
Andrew Stanley
1. It is dangerous to trust someone to give good advice when their objective is to sell you something.
2. Implement first, specify second does not spend the state's money wisely
Jim Warner
From what i watch and read until it seems that many thing are not up to date regarding today technologies...

I need to double check what I'm saying but this is my first thought...

In the other way, I was happy to find a blog/site that talk about design.

Would be better if we can have more real life example and experience
Salah CHALAL
Brilliant resource that covers many areas across the technologies we use. Different customers have varying requirements and there is always a deep dive in the areas that I need most. Covers anything from “normal” scale networks and goes right up to cloud scale. Explores the rationale behind why certain things are done in a specific manner rather than just explain what is being done.
Peter McCreesh
Great starter. The thing that I like about this is the session focusses more on why we need such architectures Vs how. This provides a wholistic view and helps think through different design options out there.
Sagar Karale
A True time value webinar
Francois DEFROCOURT
Nice start for learning data center topologies.
Engin Eser

About the instructor

Ivan did a great job in this webinar to compile all vendors' products and put together a summary with all you need to know about building data center fabric.
Krunal Shah
Danesh is a good presenter
Hassan HBAR
Very knowledgeable. Breadth of experience beyond Cumulus.
Andrew Stanley

About the materials

The voice quality was not good (mono?).
Please invest in a good microphone.

For the rest, awesome content....as always!
Best regards,
Marc Enders
Marc Enders
Excellent resource and well worth the subscription.
Peter McCreesh
One aspect that I'd love to see covered is the Datacenter physical aspect in building CLOS, traditional Core/Agg/ToR etc networks.
Sagar Karale
It would be great if you can add more information about layer 3 only data center topologies, use cases e.t.c. How multi tenancy is solved, how service integration is solved.
Also it is worth to mention that with EVPN VXLAN fabrics there is hardware limitation which limits the maximum number of switches in a single EVPN domain; number of maximum VTEP peers.
And also it would be great to mention about SRV6 usage on data centers to build overlays.
Engin Eser

Feedback from the attendees

You really save a lot of time for other people pouring all your experience and lessons learned into these webinars. A lot of time you have to ramp up in a new technology and don't have the time/experience to bring years of knowledge to bear on the matter. By cutting to the chase w/ all the media jargon, it saves a lot of time in having to discern what is marketing hype vs. what is real. I like the fact that you help us avoid the potentials of vendor lock-in. Keep up the good work!
Ahmed Aden
Great job! The information was valuable. My only recommendation would be to allocate more time for questions since this did go over the scheduled time, I had to leave and miss good information (I will watch the recording later). Thanks again for sharing your wealth of knowledge.
Shawn Mall

Tweets

@ioshints Looking forward to it. Going to join in a few minutes. Just need to grab a cup of 0xc0ffee first
@verbosemode
Just finished listening to @ioshints Datacenter Fabric webinar. HUGE amount of work on his part and incredibly well done
@abnerg
Thanks @ioshints for both the overview and highlights of the various DC fabric architectures! (webinar in http://ipspace.net)
@majornetwork
Catching up on the @ioshints Data Centre Fabric Architectures webinar #goodstuff
@jk22262
@ioshints the DCFA recording is great
@kajtzu
Just watched the 'Data Centre Fabrics' webinar.. really nice work.. wish i could get up early enough here in OZ to watch one live
@shainsingh

The Authors

Ivan PepelnjakIvan Pepelnjak, CCIE#1354 Emeritus, is an independent network architect, book author, blogger and regular speaker at industry events like Interop, RIPE and regional NOG meetings. He has been designing and implementing large-scale service provider and enterprise networks since 1990, and is currently using his expertise to help multinational enterprises and large cloud- and service providers design next-generation data center and cloud infrastructure using Software-Defined Networking (SDN) and Network Function Virtualization (NFV) approaches and technologies.

Ivan is the author of several books covering data center technologies, highly praised webinars, and dozens of data center and cloud-related technical articles published on his blog.

More about Ivan Pepelnjak

Guest Stars

Dinesh DuttDinesh Dutt has been in the networking industry for the past 20 years, most of it at Cisco Systems. Most recently, he was the Chief Scientist at Cumulus Networks, working on simplifying configuration and operations with inventions such as BGP Unnumbered and NetQ. Before Cumulus, he was a Fellow at Cisco Systems. He has been involved in enterprise and data center networking technologies, including the design of many of the ASICs that powered Cisco's mega-switches such as Cat6K and the Nexus family of switches. He also has experience in storage networking from his days at Andiamo Systems and in the design of FCoE. He is a co-author of TRILL and VxLAN and has filed for over 40 patents.

Dan Backman
Dan Backman has been in the networking industry since 1996 and currently works as a Solutions Architect at Plexxi, where he focuses on datacenter deployments as well as network integration and scaling. Dan specializes in designing and deploying large, inter-datacenter MPLS networks for large financials and utilities, and he holds JNCIEs for both service provider and enterprise. Dan joined Plexxi from Juniper Networks where he held roles as a consulting engineer, systems engineer, and most recently a technology strategist.

More about Dan Backman...

Roger Lapuh
Roger Lapuh holds a role as a Senior Product Line Manager and Architect for Ethernet Switching at Avaya. In this global position he is responsible for strategic Enterprise Campus and Data Center features for Avaya’s data portfolio. In recent years Roger focused on next generation network virtualization based on IEEE 802.1aq (Shortest Path Bridging), where Roger has co-authored, with other exponents of the industry, an IP/SPB draft as well as several patent submissions that significantly simplify network virtualization. In addition one of his key achievements has been the introduction of the Avaya (formerly Nortel) Multi-Chassis LAG solution (SMLT/RSMLT) in 2001. He is a holder of several patents in this area.

Pete LumbisPete Lumbis, CCIE #28677, CCDE 2012::3, is a systems engineer and datacenter architect at Cumulus Networks. Pete helps customers build and design networks with a focus on scale, flexibility and automation. Prior to Cumulus Pete was an escalation engineer for routing protocols, including BGP, multicast and MPLS in the Cisco TAC.

Lukas KrattigerLukas Krattiger, CCIE No. 21921 (Routing and Switching/Data Center), is a Principal Technical Marketing Engineer (PTME) with more than 18 years of experience in Data Center-, Internet- and Application-Networks. Within the Cisco Enterprise Infrastructure and Solutions Group (EISG), he specializes in Data Center switching architectures and solutions across platforms. Lukas is a double-CCIE (R&S, Data Center) with several other industry certifications and has participated in various technology leadership and advisory groups.

More about Lukas Krattiger...

Mark NowellMark Nowell is a Fellow at Cisco Systems Inc. He is currently focused on next-generation interconnect technology innovations to meet Cisco's needs. Throughout the time at Cisco, he had lead numerous product development teams including Architecture, ASIC, Board Design, Mechanical, Thermal, and Optical. His core technical background is within the optics area. He is also active in leadership roles in a number of industry forums. You can find more about Mark on his LinkedIn profile.

More about Mark…

Pete LumbisPete Lumbis, CCIE #28677, CCDE 2012::3, is a systems engineer and datacenter architect at Cumulus Networks. Pete helps customers build and design networks with a focus on scale, flexibility and automation. Prior to Cumulus Pete was an escalation engineer for routing protocols, including BGP, multicast and MPLS in the Cisco TAC.