1. No category

DATA CENTER
NETWORKING
TOPOLOGIES
1
OVERVIEW
Data Center Physical Layout
 Data Center Network Topologies
 ToR vs. EoR
 Data Center Networking Issues
 Data Center Networking Requirements

2
DATA CENTER STANDARDS
ANSI/TIA-942 Telecommunications
Infrastructure Standard for Data Centers
 Published 2005 – available through TIA at
www.tiaonline.org
 ANSI/NECA/BICSI-002 Data Center Design and
Implementation Best Practices
 complements TIA-942 –2007

3
PURPOSE OF TIA-942
Encourage early participation of telecom
designers in data center design process
 Fill a void by providing standards for planning of
data centers, computer rooms, serverrooms, and
similar spaces.
 The standard encompasses much more thanjust
telecommunications infrastructure.
 Close to half of the technical content deals with
facility specifications.

4
PURPOSE OF TIA-942

Define a standard telecommunications
infrastructure for data centers
Structured cabling system for data centers using
standardized architecture and media
 Accommodates a wide range of applications (LAN,
WAN, SAN, channels, consoles)
 Accommodates current and future protocols (e.g., 10+
GbE (Giga bit Ethernet) )
 Replaces unstructured point-to-point cabling that
uses different cabling for different applications

5
DESIGN ELEMENTS
1.
2.
3.
Cabling Design
Facility Design
Network Design
6
CABLING AND FACILITY DESIGN

Cabling Design:





Copper and fiber cabling performance
Connectors, cables, distribution hardware
Cabling distances
Space management
Facility Design:





Data center sizing
Power distribution methodologies
Pathways and spaces
HVAC, security, operations, and
administration.
Flexibility, scalability, reliability and space
management
7
NETWORK DESIGN:
Support of legacy systems
 Enable rapid deployment of new and emerging
technologies such as 10 GbE and 10+ GbE
 copper and fiber applications.

8
GOOGLE DATA CENTER
Take a walk through a Google data center
 http://www.google.com/about/datacenters/inside/s
treetview/

9
COOLING PLANT
10
MODULAR DATA CENTERS
Called Ice cube
 Small: < 1 MW, 4 racks per unit
 Medium: 1-4 MW, 10 racks per unit
 Large: > 4 MW, 20 racks per unit
 Built-in cooling,
 high PUE (power usage effectiveness) 1.02 PUE
= Power In/Power Used. It means cooling use 2
percent of power. Not efficiency of 0.98!
 Rapid deployment

11
CONTAINERIZED DATA CENTER

All companies like IBM and cisco make it like
this
12
UNSTRUCTURED CABLING
13
STRUCTURED CABLING
14
EQUIPMENT CABINETS
Three Layers: 1. Bottom: Signaling (Ethernet),2.
Middle: Power and 3. Top: Fiber
 Minimize patching between cabinets and racks

15
DATA CENTER PHYSICAL LAYOUT
16
ANSI/TIA-942-2005 STANDARD
17
ANSI/TIA-942-2005 STANDARD








Computer Room: Main servers
Entrance Room: Data Center to external cabling
Cross-Connect: Enables termination of cables
Main Distribution Area (MDA): Main cross connect. Central
Point of Structured Cabling. Core network devices
Horizontal Distribution Area (HDA): Connections to active
equipment.
Equipment Distribution Area (EDA): Active
Servers+Switches.
Zone Distribution Area (ZDA): Optionally between HDA and
EDA. Rack, cabinet, or under floor enclosure that houses a
zone outlet (ZO) or consolidation point (CP)
Backbone Cabling: Connections between MDA, HDA, and
Entrance room
18
ZONE DISTRIBUTION AREA
19
DATA CENTER NETWORK TOPOLOGIES

Three levels of switches: Core, Aggregation, Access
20
DATA CENTER NETWORKS
20-40 servers per rack
 Each server connected to 2 access switches with 1
Gbps (10 Gbps becoming common)
 Access switches connect to 2 aggregation
switches
 Aggregation switches connect to 2 core routers
 Aggregation layer is the transition point between
L2-switched access layer and l3-routed core layer
 Low Latency: In high-frequency trading market,
a few microseconds make a big difference.
 Cut-through switching and low-latency
specifications.

21
DATA CENTER NETWORKS (CONT)
Edge routers manage traffic between aggregation
routers and in/out of data center
 All switches below each pair of aggregation
switches form a single layer-2 domain
 Each Layer 2 domain typically limited to a few
hundred servers to limit broadcast
 Most traffic is internal to the data center.
 Network is the bottleneck. Uplinks utilization of
80% is common.
 Most of the flows are small. Mode = 100 MB.
DFS (Distributed File System) uses 100 MB
chunks.

22
SWITCH LOCATIONS
23
TOR VS EOR


ToR:
Advantages:
Easier cabling
 If rack is not fully populated


unused ToR ports
DisAdvantages:
If rack traffic demand is high, difficult to add more ports
 Upgrading (1G to 10G) requires complete Rack upgrade



EoR:
Disadvantages


Longer cables
Advantages:
Severs can be place in any rack
 Ports can easily added, upgraded

24
HIERARCHICAL NETWORK DESIGN
All servers require application delivery services
for security (VPN, Intrusion detection, firewall),
performance (load balancer), networking (DNS,
DHCP(Dynamic Host Control Protocol), NTP
(Network Time Protocol), FTP, RADIUS (Remote
Access Dial-In User Service server to perform
authentication ), Database services (SQL)
 Stateful devices (firewalls) on Aggregation layer

25
DATA CENTER ACCESS LAYER DESIGN

1.
2.
3.
4.
4 Possibilities:
Looped Triangle
Looped Square
Loop Free U
Looped Free Inverted U
26
ACCESS AGGREGATION CONNECTIONS
27
CISCO USES ICONS TO REPRESENT
DIFFERENT NETWORKING DEVICES
28
DATA CENTER NETWORKING ISSUES
29
DATA CENTER NETWORKING ISSUES
(CONT)
Under-utilization. Even when multiple paths
exist only one is used.
 ECMP (Equal Cost Multipath) is used by routers
to spread traffic to next hops using a hash
function. However, only 2 paths exist.

30
DCN REQUIREMENTS
Needs to be Scalable, Secure, Shared,
Standardized, and Simplified (5 S's)
 Converged Infrastructure: Servers, storage, and
network have to work together
 Workload Mobility: Large L2 domains required
for VM mobility
 East-West Traffic: Significant server-to-server
traffic as compared to server to user. One
Facebook request required 88 cache looks, 35
database lookups, 392 backend RPC calls.
Intranet traffic 935X the http request/response
 Storage traffic on Ethernet: Congestion
management on Ethernet

31
4-POST ARCHITECTURE AT FACEBOOK
Each rack contains a rack switch (RSW) with up
to forty-four 10G downlinks and four or eight 10G
uplinks (typically 10:1 oversubscription) one to
each cluster switch (CSW)
 A cluster is a group of four CSWs and the
corresponding server racks and RSWs

32
4-POST ARCHITECTURE AT FACEBOOK
(CONT)







Each CSW has four 40G uplinks (10G×4), one to each of
four “FatCat” aggregation switches (typically 4:1
oversubscription).
The four CSWs in each cluster are connected in an 80G
protection ring (10G×8) and the four FC switches are
connected in a 160G protection ring (10G×16)
network failures used to be one of the primary causes of
service outages. The additional redundancy in 4-post has
made such outages rare.
traffic that needed to cross between clusters used to
traverse expensive router links. The addition of the FC
tier greatly reduced the traffic through such links.
All tiers are switched and not routers
Ref: N. Farringon and A. Andreyev, “Facebook’s Data Center Network Architecture,”
2013 IEEE Optical Interconnect Conference,
33
http://nathanfarrington.com/papers/facebook-oic13.pdf
MAIN DISADVANTAGES OF
1.
2.
3.
4.
5.
4-POST
A CSW failure reduces intra-cluster capacity to
75%
The cluster size is dictated by the size of the
CSW
Large switches are produced in smaller volumes
from fewer manufacturers
Large switches are often oversubscribed
internally, meaning that not all of the ports can
be used simultaneously
Large switches are often very proprietary. This
can lead to months and years between bug fixes
34
CLOS NETWORKS
35
FAT-TREE
Fat-Tree networks were proposed by Charles E.
Leiserson in 1985. Such network is a tree, and
processors are connected to the bottom layer.
 The distinctive feature of a fat-tree is that for any
switch, the number of links going down to its
siblings is equal to the number of links going up
to its parent in the upper level
 Therefore, the links get “fatter” towards the top
of the tree, and switch in the root of the tree has
most links compared to any other switch below it:

36
FOR ENTERPRISE NETWORKS




However, for enterprise networks that connect
servers, commodity (off-the-shelf) switches are used,
and they have a fixed number of ports. Hence, the
design of fat-tree, where the number of ports varies
from switch to switch, is not very usable.
Therefore, alternative topologies were proposed that
can efficiently utilize existing switches with their
fixed number of ports.
There is a controversy whether such topologies should
be called fat-trees, or rather “(folded) Clos networks”,
or anything else.
However, the term fat-tree is widely used to describe
such topologies.
37
TWO-LEVEL FAT-TREE NETWORK.
38
SUMMARY
1.
2.
3.
4.
5.
Modular data centers can be used for easy
assembly and scaling
Three tiers: Access, Aggregation, Core
Application delivery controllers between
Aggregation and core
Need large L2 domains
Fat-tree topology is sometimes used to improve
performance and reliability
39