Scaling Data Center Networks

Slides:



Advertisements
Similar presentations
© Antônio M. Alberti 2011 Host Identification and Location Decoupling: A Comparison of Approaches Bruno Magalhães Martins Antônio Marcos Alberti.
Advertisements

Software Architecture Design Chapter 12 Part of Design Analysis Designing Concurrent, Distributed, and Real-Time Applications with UML Hassan Gomaa (2001)
Copyright © 2004 Juniper Networks, Inc. Proprietary and Confidentialwww.juniper.net 1 E-VPN and Data Center R. Aggarwal
Why do current IP semantics cause scaling issues? −Today, “addressing follows topology,” which limits route aggregation compactness −Overloaded IP address.
A Unified LISP Mapping Database for L2 and L3 Network Virtualization Overlays Draft-hertoghs-nvo3-lisp-unfied- control-plane Yves Hertoghs.
Hierarchical Routing Architecture Introduction draft-xu-rrg-hra-00.txt Routing Research Group Xiaohu XU
IEEE OmniRAN for Heterogeneous Networks 24 July 2012 Roger Marks Juan Carlos Zuniga Charlie Perkins 1.
Network Virtualization Overlay Control Protocol Requirements draft-kreeger-nvo3-overlay-cp-00 Lawrence Kreeger, Dinesh Dutt, Thomas Narten, David Black,
PortLand: A Scalable Fault-Tolerant Layer 2 Data Center Network Fabric. Presented by: Vinuthna Nalluri Shiva Srivastava.
IPv4 and IPv6 Mobility Support Using MPLS and MP-BGP draft-berzin-malis-mpls-mobility-00 Oleg Berzin, Andy Malis {oleg.berzin,
Network Overlay Framework Draft-lasserre-nvo3-framework-01.
Gap Analysis of Simplified Use of Policy Abstractions (SUPA) Presenter: Jun Bi draft-bi-supa-gap-analysis-02 IETF 92 SUPA BoF Dallas, TX March 23, 2015.
BGP L3VPN Virtual PE draft-fang-l3vpn-virtual-pe-01
MPLS And The Data Center Adrian Farrel Old Dog Consulting / Juniper Networks
Model-based Programmable Networks
Lucy Yong Susan Hares September 20, 2012 Boston
Draft-bitar-nvo3-vpn-applicability-00.txt Page - 1 Cloud Networking: Framework and VPN Applicability draft-bitar-nvo3-vpn-applicability-00.txt Nabil Bitar.
VPN4DC Discussion VPN4DC Team Taipei, Taiwan.
1 Mobility Support by the Common API for Transparent Hybrid Multicast draft-irtf-samrg-common-api-03 Project Matthias Wählisch,
Vic Liu Liang Xia Zu Qiang Speaker: Vic Liu China Mobile Network as a Service Architecture draft-liu-nvo3-naas-arch-01.
BCP for ARP/ND Scaling for Large Data Centers
Distributed Information Systems. Motivation ● To understand the problems that Web services try to solve it is helpful to understand how distributed information.
BGP L3VPN Virtual CE draft-fang-l3vpn-virtual-ce-01 Luyuan Fang Cisco John Evans Cisco David Ward Cisco Rex Fernando Cisco John Mullooly Cisco Ning So.
Copyright 2009, USC/ISI. All rights reserved. 11/30/2015 5:20 AM 1 Recursive Networks Joe Touch USC/ISI With: Yu-Shun Wang Lars Eggert Venkata Pingali.
MPLS-TP INTER-OP: WHAT, WHY, AND HOW? General Objectives for MPLS-TP Inter-Op Test Program at UNH-IOL.
Mar del Plata, Argentina, 31 Aug – 1 Sep 2009 ITU-T Kaleidoscope 2009 Innovations for Digital Inclusion Ved P. Kafle, Hideki Otsuki, and Masugi Inoue National.
Architecture View Models A model is a complete, simplified description of a system from a particular perspective or viewpoint. There is no single view.
VXLAN Nexus 9000 Essentials for the Data Center Karim Afifi
Network Virtualization Overlay Control Protocol Requirements draft-kreeger-nvo3-overlay-cp Lawrence Kreeger, Dinesh Dutt, Thomas Narten, David Black, Murari.
Evaluate container lifecycle support in TOSCA TOSCA – 174 Adhoc TC.
Why Fabric? 1 Complicated technology/vendor/device specific provisioning for networks, especially heterogeneous network DC Network – STP, TRILL, SPB, VXLAN,
J. Halpern (Ericsson), C. Pignataro (Cisco)
IP/MPLS VPN Protocol GAP Analysis For NVO3 draft-hy-nvo3-vpn-protocol-gap-analysis-02 Lucy Yong Susan Hares March 2013 Orlando FL.
Network Virtualization Ben Pfaff Nicira Networks, Inc.
2V0-641 Dumps VMware Certified Professional 6 - Network Virtualization Beta Exam PDF File Download :
MPLS Virtual Private Networks (VPNs)
TRILL DataCenter/Campus/PBB Inter-connect over IP core with BGP
Zhenbin Li, Kai Lu Huawei Technologies IETF 98, Chicago, USA
GENUS Virtualisation Service for GÉANT and European NRENs
Towards an Evolvable Internet Architecture
Containers: The new network endpoint
Container-based Operating System Virtualization: A scalable, High-performance Alternative to Hypervisors Stephen Soltesz, Herbert Potzl, Marc E. Fiuczynski,
Operating a Secure Network
Dave Allan Requirements and Framework for Unified MPLS Sub-Network Interconnection draft-allan-unified-mpls-req-frmwk-00 Dave.
Connecting LANs, Backbone Networks
Hierarchical Fabric Designs
Framework of Network Virtualization for Future Networks
NSI Topology Thoughts on how topology fits into the NSI architecture
1.
Overview of SDN Controller Design
Service Provider Requirements for Ethernet Control with GMPLS
A Survey of Routing Techniques for Mobile Communication Networks
Sensor Data Collection Through Unmanned Aircraft Gateways
Debashish Purkayastha, Dirk Trossen, Akbar Rahman
Kireeti Kompella Juniper Networks
T Research Seminar on Datacommunications Software
ModelNet: A Large-Scale Network Emulator for Wireless Networks Priya Mahadevan, Ken Yocum, and Amin Vahdat Duke University, Goal:
See your OpenStack Network Like Never Before
FARA: Reorganizing the Addressing Architecture
EVPN a very short introduction
An Introduction to Software Architecture
IEEE 802 Scope of OmniRAN Abstract
Attilla de Groot | Sr. Systems Engineer, HCIE #3494 | Cumulus Networks
10th International Conference on Telecommunication, ICT’2003,
MICROSOFT NETWORK VIRTUALIZATION
Host and Small Network Relaying Howard C. Berkowitz
Applicability of EVPN to NVO3 Networks
Applying CIM to SD-WAN Weiqiang Cheng, Feng Yang(CMCC)
Tim Strakh CEO, IEOFIT CCIE RS, CCIE Sec CCIE Voice, CCIE DC
Presentation transcript:

Scaling Data Center Networks Ignas Bagdonas ignas.bagdonas@cisco.com

Scaling Data Center Networks DC players: the Application and the Network. Different interdependent entities and layers Scalability is a broad topic. Different scaling problems for 10 hosts and 10M hosts The topic is rapidly evolving and may be controversial There are multiple concepts in different stages of development in various SDOs. Assuming some foundation knowledge about data center technologies This is not a technology tutorial and does not cover deep technology details It does not cover every possible scalability aspect related to DC networking

DC Network Architecture Context A very abstracted view from high above End System Hypervisor Top of Rack End of Row Edge Gateway Have we been here before? H H VM VS VM ToR ToR VM VS VM VM EoR H VM VM VS VM VM GW

Scalability Requirements It has to work. L2 and L3 identifier scope and rate of growth L3 to L2 mapping Identifiers required for data plane isolation Amount of logical infrastructure entities OAM scalability External connectivity separation and isolation

Address Resolution and Mapping Guest VMs cannot be easily modified Edge gateway needs to have all L3/L2 bindings EoR and ToR nodes build dataplane driven forwarding topology Host addressing is stable Possible solution - L2/L3 mapping Possible solution - shim interceptor layer and lookup directory Both options lead to definition of Network Controller entity Both options interwork with orchestration mechanisms closely Security aspect – any to any reachability is not required

Mobility L2/L3 mapping works for static topology. VMs can move and the pattern is hardly predictable from network point of view. Routing system scale limitations VM to hypervisor to network signalling ToR node state estimation This leads to definition of Application Controller entity Application and Network Controllers must interwork

Overlays and Underlays Partition the network by building another layer and isolate as close to VM as possible, yet not requiring changes to VM. VN-ID: another layer requires another identifier namespace, abstracted away from data plane Overlay Gateway function (External, Intra-DC, Inter-DC) Capsulation points: Hypervisor or ToR? Host signalling Edge signalling and external connectivity interworking

Overlay Architectures VM VS OV ToR Overlay capsulation points can be on hypervisor or ToR or a standalone function H VM VS OV ToR EoR H VM VS ToR OGW GW

Topology Control Overlay tends to hide underlay capabilities Multidestination forwarding constructs do not help scaling Distributed control plane is required but not enough Feedback and interworking with application controller allows for more effective network topology control

Overlay Technology Options Multiple options having different data plane realization, addressing namespace, and control plane options PBB, SPB(M), TRILL VXLAN, NVGRE, OTV EVPN LISP STT MPLS family Pick any type of data plane representation, L2/L3 address namespace, control plane complexity, ease of OAM integration, end system interworking, and architectural brokenness.

Operational Requirements Interwork with orchestration Identifier manageability and scope Service and network discovery Mapping/directory system OAM Security

Scalability Revisited Overlay function in hypervisor First hop data plane ToR state Gateway state Edge state

Summary Adding another layer always solves the problem? Controllers and dynamic control plane need to interwork. Many aspects of scalability have been studied and implemented in SP networks – DC network architecture is starting to move into similar direction.