0% found this document useful (0 votes)
285 views19 pages

Huawei Cloud Campus Networks: Agile & SDN-Ready

describe SDN evolution and bet practice
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
285 views19 pages

Huawei Cloud Campus Networks: Agile & SDN-Ready

describe SDN evolution and bet practice
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 19

Huawei Cloud Campus Networks

Agile & SDN-Ready


Steven Harrison, Senior Technical Architect, West Europe
Huawei Agile Education Infrastructure

3rd
Location Mobile Party
High Density Wi-Fi Agile Switch Router FW Controller NMS
Server APP

Design Deployment Operation Maintenance

ICT Infrastructure

… Service Enablement for End Users


(Students, Teachers, media, employees ..)

2
Evolution from Traditional Campus to Cloud Campus Based on SDN
Traditional Campus Agile Campus

WAN/Internet WAN/Internet
Agile Controller
Campus Controller (Agile Brain)

Centralized SDN control: network-


wide collaboration, unified network
Firewall single-point defense resources scheduling

L3
Independent Agile Switches (Agile Body)
ACs Security resource center L2 Intelligent awareness and execution:
Wireless network
authentication point
users/applications/quality/fault; centralized
Wireless network
On-demand allocation authentication/policy control of wired and
policy control point L3 Access/Aggregation of security resources wireless network users
switch
L2 Wired network authentication Zero-Configuration Access
point
Wired network policy control Switches (easy access)
point Simplify deployment and O&M
Distributed policy deployment of
Reduce requirements on
the entire network
Access/Aggregation switches

Central AP

Tunnel
Policy delivery

3
Agile Campus Network Solution
Service System
O&M Efficiency
IT Depart

Digital
library
Remote
education
Mobile
learning
Virtual
experiment 30 days

Internet
30 minutes
Agile Campus Network(SDN ) Agile Network
High Speed Wi-fi & Wired Network

Multimedia Office Library Dormitory

Buildings Branch

4
Smart Classroom

HDMI RS232/485
Cloud
classroom
Campus Multimedia Security alarm

Audio USB

Campus
network Whiteboard

Wi-Fi ZigBee
AR160 AR160

Mobile terminal Energy Detection

 Faults on interfaces and devices can be


accurately detected.
 Built-in IoT capabilities allow classroom
environment adjustment based on teaching
needs, improving the teaching efficiency.
Classroom  Various IoT interfaces allow a variety of devices
to access the network.

5
Distributed Wi-Fi Architecture: Improves User Experience

Distributed Wi-Fi Architecture


Cloud AC

Switch
Central AP
PoE

... ...
RRU RRU RRU RRU RRU RRU RRU RRU RRU: Remote Radio Unit

Wi-Fi Performance Improved by Management Nodes Reduced by


Pervasive Indoor Coverage Zero Roaming
20% 90%+

Traditional AP Central AP
RRU Cloud AC
Air interface
Roaming switchover Load balancing
management

Radio control Spectrum analysis Association control


Node
Central AP
RRU ... ...

Air interface RRU RRU RRU


management
Functional modules reduced by Roaming domain
Wall plate- Wall-mounted ... ...
Radio control 30%
mounted
RRU RRU RRU

6
High-Density Boost Technology for Multi-user Access

Traditional Wi-Fi Network Huawei High-Density Wi-Fi Solution

Signal available, but connection Scenario-specific products


failed • Conference rooms: Deploy APs with built-in high-density smart
High-density user access causes network antennas to improve concurrent performance by 50%.
• Stadiums and lecture halls: Deploy industry's first Software-Defined
congestion, dramatically decreasing available
Radio (SDR) APs to improve concurrent performance by 35% and
network bandwidth.
reduce the number of APs to be deployed by 30%.

Professional network planning tools


• Automatic survey, planning, and implementation tool suite
• Built-in experience library and full coverage
• Wizard-based design and graphical report

Connected, but unstable


Severe network interference causes frequent,
unexpected disconnections. 20+ Successful stadiums
Network roaming leads to service interruption.
• Home stadium of Dortmund Stadium in Germany (80,000 users)
• China’s National Stadium (80,000 users)
• Home stadium of AFC Ajax in the Netherlands (50,000 users)

7
Automated Network Planning: Wi-Fi Network Requiring No Manual Management

Automated Network Planning


Conventional Solution Performance Comparison (Time)
Solution

Automated
Conventional
Network Planning
1. Network planning Solution
Solution
Before deployment
Network
planning 0 1 day

2. Parameter setting
Parameter
During deployment setting 30 minutes 6 hours
Automatic Deployment
• Automatic radio management
Deploys plug-and-play Access Points (APs) with no
network planning required. Parameter
• Automatic parameter optimization adjustment Automatic Weekly
3. Parameter adjustment Automatically retains optimal configurations for channels
and power.
After deployment • K-Best detection algorithm
Scenario: 300 APs deployed in
Directly calculates the optimal channel for APs, completing
network-wide parameter optimization in 30 minutes. an open space

8
Cloud Data Center Network Solution
 Open:
 ODL-based open-source controller architecture,

 Building an open ecosystem;

 Cloud Fabric OpenLab

HPC Big Data Social Network IOT SDN Mobility


 Flexible:
 hardware/software/hybrid VXLAN networking,
Public Cloud Private Cloud Hybrid Cloud
 centralized/distributed gateways

 Highly Efficient:
100G Cloud Fabric Security
 Fully-connected SDN, Virtualization
Integrated DCI &Disaster Recovery
High-
Elastic Virtualized Agile  Super Virtual Fabric build one virtual switch
quality Management
 On-demand and automatic deployment

 one-stop service, network, and quality management

Network automation Precise network

80%
Batch automation
measurement

Super Virtual Fabric (SVF)


One Network = One Device
9
Huawei Cloud Fabric SDN Automated Deployment Solution

Cloud OS
Cloud Stack 3rd-party OpenStack

 Application network  Logical network  Physical network topology


topology topology
• Service-driven automatic network configuration
User Web App DB
• 3-layer network visibility improves management
Agile efficiency and simplifies O&M
Controller VM
• E2E service chain orchestration improves
Service-chain deployment efficiency by 10x
orchestration VM VM

• Flexible VXLAN solutions give more choices


VXLAN+EVPN for automatic network resource allocation
Fabric vSwitch vSwitch vSwitch

Hardware VXLAN Hybrid VXLAN


• One architecture compatible with bare metal
servers, VMs, containers
ESXi Hyper-V FusionCompute KVM
Host
VM VM VM VM VM VM VM VM

10
SDN Accelerates Service Provisioning: Weeks to Minutes

Traditional Data Center Network Software-Defined Cloud Data Center

Manual network deployment Automatic network deployment


App 1 App 2

• One-click service deployment


• Different languages used on Application
administrator Policy1 Policy3

network devices and IT systems WEB1 APP1

Policy2
WEB1 APP1

Policy4

• Complex network configurations, no


Service DB1 DB1
• Administrators define and deploy
Network
automatic deployment capabilities Application App templates translated networks based on service
requirements into logical networks
requirements.
• Different interfaces and standards
used by different vendors • The AC converts the application view
Logical
Network LN1 LN2
into the logical network view, and
• Lack of automatic O&M tools
Agile
Controller
Logical templates mapped to automatically delivers configurations
physical networks
to physical networks.
• Physical and virtual networks are
vSwitch
vFW
vRouter
Physical
Network Virtual network Physical network uniformly deployed.

Service provisioning: Weeks Service provisioning: Minutes


Policy adjustment: Minutes
Policy adjustment: Days

11
E2E Service Orchestration Improves Deployment Efficiency

Drag-n-drop experience
• VAS deployment by drag-n-drop operations on GUI

Agile FW IDP LB NAT


Controller E2E, On-Demand Automated Delivery
• Automated VAS delivery including FWs, LBs, etc.
• Unified deployment of network configurations & security policies
on hardware and software FWs

Service chain

Flexible orchestration, dynamic VAS


insertion
A • Flexible service orchestration without being restricted
Terminal group
by physical networking
Web App DB
• Based on PBR and standard NSH solution(ENP)
• Open architecture interoperable with multi-vendor VAS
devices

12
SDN Realizes Network Visibility, Simplifying O&M
55/150
Used/Total
Visible service quality Network resources

Visible O&M, Simple O&M • ENP-based real flow quality measurement 0/150
Abnormal/Total
(packet loss + delay): iPCA
Network status
• Microsecond-level delay measurement:
NTP (100 ms), 1588v2 (10ns) 60% VxLAN
Network usage
• Millisecond-level packet loss measurement
WEB DB
APP

Application network topology


Visible network topologies
• Quick physical failure point location for a
Logical network topology
service failure
• Fault location time <1s Virtual network Physical network

Physical network topology

Visible service paths with VMs


• E2E connectivity detection for real service
paths
• VM-to-VM path detection time <10s

13
Hybrid VXLAN: Automatic Network Resource Allocation

Tenant Cloud Platform


Better resource utilization
Administrator Administrator
• Suitable for network reconstruction, with CE1800V used
as VTEP and old ToR switches reused
Cloud manager 3rd-party cloud manager • Flexible hybrid solution, supporting both VMs and
physical servers
Cloud platform OpenStack
More flexible network
Control VMM Controller O&M eSight • VXLAN endpoints on both software switch CE1800V and
hardware switches

Fabric
• Local switching of east-west traffic by DVR (CE1800V),
VAS group avoiding circuitous forwarding paths

VXLAN GW
• Both ESX and hyper-V platforms supported, on which

VXLAN+EVPN •
CE1800V can be installed based on VMs
Independent GUI provided by Agile Controller for
VTEP automated network service orchestration
• Interoperation with cloud platforms for cloud-network
VTEP VTEP integration
ESXi vSwitch FusionCompute
Servers VM VM VM VM VM VM VM VM VM
High performance
3rd KVM/Xen • 4K physical servers managed by a single controller node
Bare-Metal VLAN network
vSphere Hyper-V Hypervisor (Old) and 64K physical servers managed by a cluster

14
Optimized DCI optical solution
Inter-city
Server Switch DC B
High Capacity

• One pair of fibers support more than 20T


• 200G/λ;>100λ With Flex grid
Inter-city Switch Switch Intra-city
DC A DC C
DC environment friendly
WDM Bearer • AC power / Low power
Solution • Front-to-back cooling

Simple Management
Server Server

• User friendly
OptiCloud OSN 902 • Easy integration to the IT system
 2U, 6 slots, 1.2T/2.4T capacity • Automation with SDN
 200G/400G per Slot, pluggable
 19inch, Front inlet, rear outlet
WDM is the best bearer network solution for DCI

15
Huawei AgilePOL All-Optical Campus Solution

Sci-tech
zone

School
NetOpen eSight unified NMS
Hotel
ONU
Transportation 1G/10G to Desk
ODN
Electric power OLT

Surveillance

Latency reduced with fiber to the desk Simplified engineering deployment, precise fault location
• Bandwidth smoothly evolves to 10 Gbit/s with fiber to the desk. • Precise fault location, reducing maintenance difficulty.
• Flattened two-layer network reduces latency. • All-optical passive network, reducing devices by 60%, equipment
• Carrying broadband, telephone, and TV services on one fiber rooms by 80%, and power consumption by 60%.
facilitates service capacity expansion. • Simplified engineering deployment, reducing in-building horizontal
wiring by 70%.

16
POL Enables a Future-Proofed Network and Meets Bandwidth Demands

Optical access technology evolution From GPON to 10G PON


OLT platform reusing ODN network reusing ONT reusing NMS reusing
Rate
OLT ONT
OLT GPON ONT
WBF GPON ONT

10G PON
WDM GPON

GPON
OLT ONT
coupler ONT
 ODSM PON: 10G/ 10G GPON ONT
10G GPON
10G @ λ, ~100 km; WBF
ONT ONT
ONT
10 GPON ONT
 WDM PON:
32*1.25G/32*1.25G,
~100 km  Large split ratios: 1:128/256/512; a single CO covers a 40 km reach.
 NG-PON:
10G/2.5G, ~60 km From 10G PON to 40G PON
 GPON:
2.5G/1.25G, ~60 km 80 Mbit/s/user GPON λ/user WDM PON
ONT ONT
2009 2010 2015 2018 Time
ONT ONT
 Fibers: Support abundant frequency
OLT
Splitter 1:64
OLT AWG 1:64
spectrum and bandwidth resources, ONT ONT
meeting the near infinite bandwidth needs
 Different enterprises/companies/departments can use different wavelengths,
of the future.
implementing independent transmission channels and related QoS.

17
POL Allows for Multi-Service High Flexibility and Extendability
Telephone

CO
Video surveillance PC

ONT
OLT TV
Residence

ODN
Conference system

Telephone system

Surveillance system
Smart cloud
Office WLAN
ISP OTT
Door access building/Hotel
ISP' OTT Parking guidance
control intercom
Intranet Broadband

PBX

 One GPON network carries FTTH home/campus broadband services and low current system services such as residential community video surveillance,
video intercom, door access control, burglar alarm, automatic meter reading, and vehicle management.
 Huawei's AgilePOL solution uniquely provides HQoS to ensure the transmission of various services, making it the best choice for building a full-service
campus network.

18
THANK YOU

Copyright©2017 Huawei Technologies Co., Ltd. All Rights Reserved.


The information in this document may contain predictive statements including, without limitation, statements regarding the future financial and operating results, future
product portfolio, new technology, etc. There are a number of factors that could cause actual results and developments to differ materially from those expressed or
implied in the predictive statements. Therefore, such information is provided for reference purpose only and constitutes neither an offer nor an acceptance. Huawei
may change the information at any time without notice.

You might also like