Guía de Diseño Del Centro de Datos

D E S I G N G U I D E

jvega

Texto tecleado

GUIA

Table of Contents | LAN-1160-EN | Page 1

Table of ContentsCorning Cable Systems shall not be responsible for the performance of third-party productsor for any incorrect installation or installation in violation of Corning Cable Systems’ specifications and procedures.

1SECTIONChapter One:Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3

What is a Data Center? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2PoP Zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2Server Zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2Storage Zone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Chapter Two: Data Center Networking Protocols . . . . . . 4-8LAN Protocols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5SAN Protocols. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-8

Chapter Three: Fiber Type and Performance. . . . . . . . . . . 9-13OM3/OM4 Laser-Optimized 50/125 µmMultimode Fiber. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-10

Fiber vs. Copper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1010G Electronics and Cooling . . . . . . . . . . . . . . . . . . . . 10-11End Equipment Through Optical FiberDistance Capabilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Transceivers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12OM3/OM4 EMBc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Introductionto Data Centers

Chapter Four: Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-18General Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14Data Center Needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14-15Intro to TIA-942 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-16Redundancy in the Data Center . . . . . . . . . . . . . . . . 17-18

Chapter Five: Designing a Scalable Infrastructure. . . . 19-20Structured Cabling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19-20Zone Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Chapter Six: Determining the Fiber Counts . . . . . . . . . . 21-25Logical Topologies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Mapping Logical Architectures to TIA-942 . . . . . 22-24Future 40G/100G Systems . . . . . . . . . . . . . . . . . . . . . 24-25

Designing the PhysicalInfrastructure

Chapter Seven: Choosing Infrastructure Components 26-38Preterminated Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . 26Standard-Density Solutions . . . . . . . . . . . . . . . . . . . . . 27-31High-Density Solutions. . . . . . . . . . . . . . . . . . . . . . . . . 32-38

Chapter Eight:Writing a DC Request for Proposal . . . . 39-46Steps Needed to Implement an RFP . . . . . . . . . . . . . . . 39Generic Specifications for PretiumEDGE™

Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39-44Generic Specifications for Pretium EDGE1UHousing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

Generic Specifications for Pretium EDGE4UHousing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Chapter Nine: Procuring the Data Center Products . . . . . . 47Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Chapter Ten: Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

Deploying the PhysicalInfrastructure

Chapter Eleven: Polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49-52Chapter Twelve: Testing and Documentation . . . . . . . . 53-62

Cable SystemTesting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53End-to-End Attenuation Testing. . . . . . . . . . . . . . . . 53-58Application Guidelines. . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Background and Trace Interpolation . . . . . . . . . . . . . . 59Test Equipment: OTDR Analysis. . . . . . . . . . . . . . . . . . . 60Documentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61Maintenance and Troubleshooting . . . . . . . . . . . . . . . 62

Chapter Thirteen: Labeling . . . . . . . . . . . . . . . . . . . . . . . . . 63-66Choosing a LabelingMethod . . . . . . . . . . . . . . . . . . . . . 63Labeling Racks and Cabinets. . . . . . . . . . . . . . . . . . . . . . 63Labeling Patch Panels and Fiber . . . . . . . . . . . . . . . 64-66

Performance Metricsand Administration

Glossary. . . . . . . . . . . . . . 67-80

3SECTION2SECTION4SECTION

5SECTION Information and Tools

Chapter One: Overview | LAN-1160-EN | Page 2

What is a Data Center?A data center, as defined in TIA/EIA-942, Telecommun-ications Infrastructure Standard for Data Centers, is abuilding or portion of a building whose primary functionis to house a computer room and its support areas.The main functions of a data center are to centralizeand consolidate information technology (IT) resources,house network operations, facilitate e-business and toprovide uninterrupted service to mission-critical dataprocessing operations.

Data centers can be classified as either enterprise (private)data centers or co-location (co-lo)/hosting (public) datacenters. Enterprise data centers are privately owned andoperated by private corporate, institutional or governmententities. Enterprise data centers support internal datatransactions and processing, as well as Web Servicesand are supported and managed by internal IT support.Co-lo data centers are owned and operated by telcosor unregulated competitive service providers and offeroutsourced IT services. Services that data centers typicallyprovide include Internet access, application or Web hosting,

content distribution, file storage and backup, databasemanagement, fail-safe power, HVAC controls, securityand high-performance cabling infrastructure. As shownin Figure 1.1, the functional areas of the data center canbe broken down into:1. Switching

• Point of Presence (PoP) Zone• Server Area Zone

2. Storage• Storage Area Network

PoP ZoneThis area of the data center is sometimes referred toas the “meet me” room. It is typically the area wherethe service provider enables access to their networks.This area contains many routers and core switches.

Server ZoneThis area of the data center provides the front-endconnection to the database servers. This area containsmany switches and servers. The protocols used to commu-nicate in this area are 1 Gigabit and 10 Gigabit Ethernet.

SAN

SAN

Server Area

Storage Switching

Server AreaServer AreaServer Area

PoP

PoP

PoP

Figure 1.1Functional Areas of the Data Center | Drawing ZA-3580

Introduction to Data Centers 1SECTIONChapter One:Overview

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Storage ZoneThis area of the data center provides the back-endconnection to data. This area contains many types ofstorage devices. The protocols used to communicate in thisarea are Fibre Channel Ethernet and small computer systeminterface (SCSI).

Regardless of the type of data center to be implemented,there are three fundamental issues, or concerns, thatshould be addressed when evaluating each area of thedata center infrastructure:

1. Manageability2. Flexibility and Scalability3. Network Efficiency

ManageabilityEnd users are looking for a higher performance, low-profile solution for a more effective overall operationof the network. Manageability is essential; without it, thecabling infrastructure takes over the data center in a shortamount of time. To increase control over the data centerinfrastructure, structured cabling should be implemented.The key benefit of structured cabling is that the userregains control of the infrastructure rather than livingwith an unmanageable buildup of patch cords and anabundance of unidentifiable cables.

Flexibility and ScalabilityFlexibility and scalability of the cabling infrastructureallow quick and easy changes with little to no impact on theday-to-day operation of the data center, as well as reducedrisk that tomorrow’s technology will render an obsoleteinfrastructure. Scalability of the data center is essentialfor migration to higher data rates and for adding capacitywithout major disruption of operations. The initial datacenter must be designed so it can be scaled quickly andefficiently as the requirements change. To meet the require-ments and demands of the data center, the topology in thedata center, as well as the actual components used to imple-ment the topology, must be explored. Both topology andcomponents, if chosen correctly, create an effective network,save time and money, and create efficiency, manageability,flexibility and scalability in the data center.

Network EfficiencyData centers have seen significant growth in size and num-bers in the past few years and should continue to see signifi-cant growth in the future as networks continue to evolveand move toward 100 Gigabit Ethernet. Due to the consid-erable growth in data centers, there is a need to have simple,efficient cabling solutions that maximize space and facilitatereduced installation time and costs. Preterminated solutionsare often the preferred solution as they provide higher fiberdensity, reduced installation time and the ability to easilyfacilitate moves, adds and changes (MACs).

Corning Cable Systems’ preterminated optical fiber cablingsolutions streamline the process of deploying an opticalnetwork infrastructure in the data center. A modulardesign guarantees compatibility and flexibility for all opticalconnectivity and easily scales as demands dictate and require-ments change. The preterminated solutions also managefiber polarity, virtually eliminating it as a concern in networkdesign, installation or reconfiguration.

Corning Cable Systems’ newest preterminated solution,Pretium EDGE™ Solutions, provides increased system densi-ty when compared to traditional preterminated systems andoffers the highest port density in the market. Custom-engi-neered components enable simple integration into commonSAN directors and switches, while the preterminated com-ponents allow for reduced installation time and faster MACs.

A well-planned infrastructure can last 15 to 20 years andwill have to be operational through multiple generationsof system equipment and data-rate increases. The followingchapters address all of the factors to be considered for awell-designed data center cabling infrastructure.

Chapter One: Overview | LAN-1160-EN | Page 3

Chapter Two:Data Center Networking Protocols

GeneralData centers contain many network transmission protocolsfor communication between electronic equipment.Ethernet and Fibre Channel are the dominant networks,with Ethernet providing a local area network (LAN)between users and computing infrastructure while FibreChannel provides connections between servers and storageto create a storage area network (SAN). See Figure 2.1.To design a structured cabling system for a data center,the designer should understand the different protocolsthat are used in each area of the data center.

LAN ProtocolsEthernetEthernet is the most widely installed LAN data transmissiontechnology and is standardized as IEEE 802.3. Ethernet istypically used in data center backbones to transmit datapackets from the core router to the access switch to theserver network interface card (NIC). Figure 2.2 illustratesthe Ethernet frame.

Ethernet originally began as a bus-based applicationwith coaxial cable as the primary bus medium that waseventually replaced with fiber and copper twisted-pairmedia. Ethernet is now deployed in data center switchnetworks with optical connectivity in the backbone andcopper connectivity that addresses short-length equip-ment interconnects.

Data center Ethernet deployments operate at speeds of 1Gand 10G utilizing predominately OM3 and OM4 multimodeoptical fiber. Multimode fiber installations usually operateat 850 nm with VCSEL transceivers. OM3 and OM4 fiberswith 850 nm VCSEL transceivers provide significanteconomic value propositions when compared to single-modefiber and DFB/FP transceivers.

LANCore Switch

EdgeSwitch

SAN Switch

StorageServer

FCFC

Ethernet

Ethernet

OM3Copper Cable

Figure 2.1Typical Data Center Architecture Today | Drawing ZA-3468

PREAMBLE

7 OCTETS

1 OCTET

SOF

TYPE DATA FCSDESTINATION

ADDRESSSOURCE

ADDRESS

6 OCTETS 2 OCTETS 4 OCTETS

46-1500 OCTETS6 OCTETS

Figure 2.2Ethernet Frame Format | Drawing ZA-3675

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Chapter Two: Data Centers Networking Protocols | LAN-1160-EN | Page 4


The IEEE 802.3z and 802.3ae task force groups releasedstandards for Gigabit Ethernet and 10 Gigabit Ethernet in1998 and 2002, respectively. The primary 1G and 10Gphysical media dependent (PMD) variants being deployedare provided in Table 2.1.

Future industry bandwidth drivers such as video applica-tions, virtualization and I/O convergence are driving theneed for network data rates beyond 10G. In response to thatneed, the IEEE 802.3ba task force was formed to developguidance for 40G and 100G Ethernet data rates. OM3 andOM4 fibers are the only multimode fibers included in thestandard. 40/100G distances for OM3 and OM4 are 100 mand 150 m, respectively. The 40/100G standard does notinclude guidance for UTP/STP copper media.

Ethernet duplex fiber serial transmission with a directlymodulated 850 nm VCSEL has been used for data rates upto 10G. Duplex fiber serial transmission becomes impracticalat 40/100G data rates due to reliability concerns when the850 nm VCSEL is directly modulated across extreme tem-peratures in the data center. Ethernet 40/100G multimodefiber PMDs (40GBASE-SR4 and 100GBASE-SR10)uses parallel optics with OM3 and OM4 fibers to mitigatethe VCSEL reliability concern.

40G Ethernet uses four 10G channels to transmitand four 10G channels to receive while 100G Ethernetuses ten 10G channel to transmit and ten 10G channelto receive. See Figures 2.3 and 2.4.

SAN ProtocolsFibre ChannelFibre Channel is a high-performance, low latency, duplexfiber serial link application with data rates of 1 Gb/s, 2 Gb/s,4 Gb/s, 8 Gb/s, 10 Gb/s and 16 Gb/s. It provides a veryreliable form of communication that guarantees deliveryof information. The Fibre Channel T11 technical commit-tees are responsible for developing transmitting guidance.Fibre Channel is used in the data center to transmit data

from the server host bus adapter (HBA) to the SANdirector to the SAN storage. Similar to Ethernet, OM3and OM4 fibers are the dominant fibers and media typeused in the SAN network. Fibre Channel networks to date

1G:Multimode 1G: Single-mode

1000BASE-SX (OM3: 1000m, OM4: 1000m) 1000BASE-LX (SM: 10 km)

10G:Multimode 10G: Single-mode

10GBASE-SR (OM3: 300m, OM4: 550m) 10GBASE-LR (SM: 10 km)

TABLE 2.1

RxRxRxRxRxRxRxRxRxRxRxRx

TxTxTxTxTxTxTxTxTxTxTxTx

Fiber Position12 1

RxRxRxRxRxRxRxRxRxRxRxRx

TxTxTxTxTxTxTxTxTxTxTxTx

Fiber Position 121

Optical TransmitterMTP Connector

Optical ReceiverMTP® Connector

ZFigure 2.3Parallel Optics for 100G Ethernet | Drawing ZA-3300

RxRxRxRx

TxTxTxTx

TxTxTxTx

RxRxRxRx

Fiber Position12 1

Fiber Position 121

Optical Transmitter MTP Connector

Optical Receiver MTP Connector



Z

Figure 2.4Parallel Optics for 40G Ethernet | Drawing ZA-3299

have exclusively used optical media for the backbone aswell as the interconnect into the electronics. SAN FibreChannel links are being designed and deployed today tosupport migration to 16G. Maximum 16G OM3 and OM4channel distances are 100 m and 125 m, respectively. FibreChannel single-mode fiber usage is minimal in the datacenter but is exclusively used for synchronization betweenprimary and secondary data center sites. T11 activity hasrecently started to develop 32G guidance. Initial objectivesare for a duplex fiber serial transmission solution withOM3 and OM4 fibers for 70-100 m distance. Table 2.2provides the T11 Fibre Channel speed roadmap.

Fibre Channel over EthernetData centers utilize multiple networks that present opera-tional and maintenance issues as each network requires dedi-cated electronics and cabling infrastructure. As previouslydiscussed, Ethernet (LAN) and Fibre Channel (SAN) arethe typical networks in a data center. Fibre Channel’s T11technical committee and the Institute of Electrical and

Electronic Engineer’s (IEEEs) Data Center Bridgingcommittee are defining standards to converge the two intoa unified fabric with Fibre Channel over Ethernet (FCoE).

FCoE is simply a transmission method in which the FibreChannel frame is encapsulated into an Ethernet frameat the server (Figure 2.5). The server encapsulates FibreChannel frames into Ethernet frames before sendingthem over the LAN and de-encapsulates them whenFCoE frames are received. Server I/O consolidationcombines the NIC and HBA cards into a single convergednetwork adapter (CNA) which reduces server cabling andpower/cooling needs. At present, the Ethernet frame isremoved at the Ethernet edge switch to access the FibreChannel frame which is then transported to the SANdirectors. FCoE encapsulation standards activity takesplace at the Fibre Channel T11.3 committee.

Product Naming Throughput (MBps) Line Rate (GBaud) T11 Spec TechnicallyCompleted (Year)

MarketAvailability (Year)

1GFC 200 1.0625 1996 1997

2GFC 400 2.125 2000 2001

4GFC 800 4.25 2003 2005

8GFC 1600 8.5 2006 2008

16GFC 3200 14.025 2009 2011

32GFC 6400 28.05 2012 2014

64GFC 12800 57 2016 Market Demand

128GFC 25600 114 2020 Market Demand

TABLE 2.2: T11 Fibre Channel Speed RoadmapE

ther

net

Hea

der

FCS

FCoE

Hea

der

EO

F

Fibr

eC

hann

elH

eade

r

CR

C

Fibre Channel Payload

Figure 2.5Fibre Channel Payload | Drawing ZA-3673

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS


Fibre Channel is a deterministic protocol that guaranteesdelivery of information. Native Ethernet has not beendeterministic and has relied on transmission controlprotocol (TCP) to retransmit dropped frames. WithFCoE, the Ethernet transport has been required to beupdated to ensure that frames/packets are lossless withoutusing TCP/IP protocol. The new enhanced Ethernetstandard is called converged enhanced Ethernet (CEE).CEE standards activity takes place at the IEEE 802.1Data Center Bridging working groups.

Table 2.3 provides the Fibre Channel Industry Association(FCIA) FCoE speed roadmap. Where 10G FCoE utilizesserial duplex fiber transmission, 40/100G FCoE speedswill require parallel optics. Data centers should install12-fiber MPO backbone cables with OM3 or OM4 fibertoday that can be used for 10G FCoE and to provide aneffective migration path to emerging parallel optics thatrequire an MPO interface into the switch electronics andthe server (Figure 2.6).

First generation FCoE implementation will focus on theedge switch and server. Ethernet OM3 or OM4 fiberoptical uplinks will be received into the FCoE enabled edgeswitch and then interconnected to the server CNA. Insteadof copper UTP interconnects, SFP+ direct attached twinaxialcopper cable is now used as the media with significantlylower power and latency performance. The twinax coppercable will be used for distances up to 7-10 m. Beyond thatdistance, low-cost, ultra-short-reach (USR) SFP+ modulesand OM3 or OM4 optical fiber will be used. The encapsulat-ed Fibre Channel frame is returned to the edge switch where

the Ethernet frame is removed to access the Fibre Channelframe. The Fibre Channel frame is then transmitted to theSAN network. See Figure 2.6. This architecture solutionreduces the server interconnect cabling and adapter cardnumber by at least 50 percent.

Second generation FCoE deployments are expectedto use FCoE enabled core switches and edge switches.This architecture will continue to use basic Ethernetoptical uplinks from the core switch to the edge switch andSFP+ twinax interconnects into the server. The differenceoccurs when the FCoE frame is transmitted back through

TABLE 2.3: T11 Fibre Channel Speed Roadmap

Product Naming Throughput (MBps) Equivalent LineRate (GBaud)

T11 Spec TechnicallyCompleted (Year)

MarketAvailability (Year)

10GFCoE 2400 10.3125 2008 2009

40GFCoE 9600 41.225 TBD Market Demand

100GFCoE 24000 103.125 TBD Market Demand

LANCore Switch

SAN Switch

StorageServer

FCFC

Ethernet

FCoE

OM3SFP+ Twinax

FCoEEdge Switch

Figure 2.6First Generation FCoE Architecture | Drawing ZA-3469


the edge switch to the core switch over the same opticalfiber previously used as the uplink to the server. At thecore switch, the FCoE frame is forwarded to the SANdirector where the Ethernet frame is removed and theFibre Channel frame is then transmitted to the storagedevices. This architecture solution reduces the serverinterconnect cabling and adapter card number byat least 50 percent and eliminates the Fibre ChannelHBA to SAN optical fiber trunk cable. See Figure 2.7.

Third generation FCoE architecture mirrors the secondgeneration with the exception that the core switch nowforwards the FCoE frame directly to storage wherethe Fibre Channel frame is accessed. This architecturesolution reduces the server interconnect cabling andadapter card number by at least 50 percent, eliminatesthe Fibre Channel HBA to SAN optical fiber trunk cableand eliminates the core switch to SAN director fibertrunk cable. See Figure 2.8.

The FCIA has adopted specific guidance relative to thecabling physical layer. Optical connectivity shall be inaccordance with IEEE 802.3ae (10GBASE-SR) utilizingOM3 or OM4 optical fiber. In addition, new installsare recommended to be = < 100 m to be compatible withemerging 40/100G Ethernet and 16/32G Fibre Channel.The SFP+ is the preferred electronic interface for copperand optical cable. This eliminates use of 10GBASE-Tcopper UTP/STP cable.

FCoE offers a data center unified fabric solution thatsimplifies operational and maintenance of the cablinginfrastructure. FCoE facilitates utilization of low-costEthernet electronics and OM3/OM4 optical connectivityto support 10/40/100G data rates.

FCoECore Switch

Storage

Server

FCoE

FCoE

OM3

SFP+ Twinax

FCoE

FC

SANSwitch

FCoEEdge Switch

Figure 2.7Second Generation FCoE Architecture | Drawing ZA-3470

FCoECore Switch

Server

FCoE

FCoE

OM3SFP+ Twinax

FCoE

StorageFCoE

Edge Switch

Figure 2.8Third Generation FCoE Architecture | Drawing ZA-3471

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS


Chapter Three:Fiber Type and Performance

As fiber becomes more widely deployed in the data center,a system designer should evaluate all the various grades ofmultimode fiber optic cable to ensure the data center willsupport current and future data rates. As data rates andthe physical size of data centers increase, the need fordesigning a bandwidth and link-length scalable networkis more important then ever. The purpose of this chapteris to familiarize the reader with OM3 and OM4 fiber typesand performance requirements needed to support localarea network (LAN) and storage area network (SAN)applications commonly used in data centers.

OM3/OM4 Laser-Optimized50/125 µmMultimode FiberData center LAN and SAN networks should be designedto support legacy applications as well as emerging high-data-rate applications. The emergence of high-data-ratesystems such as 10, 40 and 100 Gigabit Ethernet and 8 and16 Gigabit Fibre Channel has resulted in OM3 and OM4multimode fibers being the dominant optical fiber typesdeployed in the data center.

The TIA-492AAAC OM3 detailed fiber standard wasreleased in March 2002, and the TIA-492AAAD OM4detailed fiber standard was released in August 2009. Thefibers are optimized for laser-based 850 nm operation and

include a minimum 2000 MHz•km effective modal band-width (EMB) for OM3 and 4700 MHz•km EMB for OM4.The OM multimode fiber nomenclature originated in theISO/IEC-11801, second edition standard and has beenadopted into TIA standards such as TIA-568, Rev C.3.In addition to OM3 and OM4, OM1 and OM2 designationsare included for standard 62.5 µm and 50 µm multimodefibers, respectively. See Table 3.1.

Data center high data rates in conjunction with the desiredapplication distances support OM3 and OM4 as the defaultchoice fiber types. The small core size of 50/125 µm fiberyields an inherent higher bandwidth capability than othermultimode fibers such as OM1 fiber. Tables 3.2 and 3.3provide OM3 and OM4 fibers distance capabilities forEthernet and Fibre Channel data rates.

Corning Cable Systems strongly recommends OM3 andOM4 fibers for the data center. When compared to OM1and OM2 multimode fibers, OM3/OM4 fibers havethe highest 850 nm bandwidth to accommodate longerdistances, provide more system budget margin and supportmigration to higher data rates such as 16/40/100G.

Optical FiberCable Type Fiber Reference Wavelength

Overfilled ModalBandwidth-LengthProduct (MHz•km)

Effective ModalBandwidth-LengthProduct (MHz•km)

62.5/125 µmmultimode (OM1)

TIA-492AAAA-AIEC 60793-2-10Type A1b

8501300

200500

Not RequiredNot Required

50/125 µmmultimode (OM2)

TIA-492AAABIEC 60793-2-10Type A1a.1

8501300

500500

Not RequiredNot Required

850 µmlaser-optimized50/125 µm (OM3)

TIA-492AAAC-AIEC 60793-2-10Type A1a.2

8501300

1500500

2000Not Required

850 µmlaser-optimized50/125 µm (OM4)

TIA-492AAADIEC 60792-2-10Type A1a.3

8501300

3500500

4700Not Required

TABLE 3.1

Chapter Three: Fiber Type and Performance | LAN-1160-EN | Page 9

Expectation is that implementing an OM3/OM4 physicallayer solution should provide a 10-15 year service lifewithout recabling.

Cable, connectors, hardware and electronics are now readilyavailable to support usage of these 50 µm fibers. The techni-cal and commercial community has recognized the benefitsof OM3/OM4 as the fibers have been adopted into IEEE40/100G and Fibre Channel 4/8/16G transmission stan-dards as well as the TIA-568-G3 structured cabling andconnectivity standards. The 850 nm wavelength now offersand will continue to offer the most economical solutionfor data center applications based on electronic costs.The data rate scalability of OM3 and OM4 fibers providesthe ultimate media solution for data center managers toensure their structured wiring systems support legacy aswell as future application needs.

Fiber vs. CopperA well-planned structured cabling system in the data centerwill support both the applications of today as well as thefuture. Corning Cable Systems’ data center solutions dojust that, allowing today’s systems to grow gracefully asrequirements change without concern of obsolescence.Fiber is the most attractive medium for structured cablingbecause of its ability to support the widest range of applica-tions at the fastest speeds for the longest distances.Additionally, fiber has a number of intrinsic advantagesbeneficial to any application at any speed. Fiber is immuneto electromagnetic interference (EMI) and radio frequencyinterference (RFI), therefore its signals cannot be corruptedby external interference. Just as it is immune to EMI from

outside sources, fiber produces no electronic emissions,therefore it is not a concern of the Federal CommunicationsCommission (FCC) or European emissions regulations.Cross-talk does not occur in fiber systems and there are noshared sheath issues as with multipair unshielded twisted-pair(UTP) copper cables. Also, standards activity has shownevidence of alien cross-talk between UTP copper cables thatcannot be corrected by electronic digital signal processing(DSP). Because all-dielectric cables, as well as the newdielectric armored cables, can be used, grounding concernscan be eliminated and lightning effects dramatically reduced.Optical fibers are virtually impossible to tap, making it themost secure media type. Most importantly, optical bandwidthcannot be adversely affected by installation conditions.Compare this to the copper system impairments that aninstaller can impact.

10G Electronics and Cooling –The Optical Advantage10G optical switch electronics and server adapter cardsrequire less power to operate compared to 10G UTP cop-per. The high insertion loss of copper cables at the extendedfrequency range needed to support 10G and the requiredelectronic digital signal processing (DSP) noise-reductioncircuitry means that energy consumption will inevitably behigher than that of low-loss fiber interconnects. 10GBASE-SR SFP+ optical transceivers consume a maximum of 1.0watt (typical 0.5 watt) per port compared to 6-8 watts perport for a 10GBASE-T copper switch. SFP+ chassis linecards are intended to support up to 48-64 ports, while10GBASE-T cards are expected to have 8-16 ports.10GBASE-SR server adapter cards typically use less


1G 10G 40G 100G

OM3 1000 300 100 100

OM4 1000 550 150 150

TABLE 3.2: 850 nm Ethernet Distance (m)

4G 8G 16G

OM3 380 150 100

OM4 480 190 125

TABLE 3.3: 850 nm Fibre Channel Distance (m)

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

than nine watts to service up to 300 m, while announced10GBASE-T cards use 24 watts to service up to 100 m.Experts have stated that 10GBASE-T over CAT 6A orCAT 7 twisted-pair can extend up to 100 m, but powerrequirements hinder its cost-effectiveness. A 10G opticalsystem requires far fewer switches and line cards for equiva-lent bandwidth capability of a 10G copper system. Fewerswitches and line cards translate into less energy consump-tion for electronics and cooling to minimize operationalexpenses and support environmental initiatives. See Figure3.1. One optical 48-port line card equals three 16-port linecards. As with the 10G copper switches, the 10G copperserver adapter card’s high power consumption and coolingneeds result in a higher operational expense. The industry10GBASE-T expectation is that three to four watts perport will be the lowest achievable power consumption.

High fiber density, combined with the small diameter ofoptical cable, maximizes raised floor pathways and spaceutilization for routing and cooling. Optical cables also offersuperior pathway usage when routed in aerial cable trays.A 0.7-inch diameter optical cable would contain 216 fibersto support 108 10G optical circuits. The 108 copper cablesrequired to provide equivalent capability would have a5-inch bundle diameter. The 10G twisted-pair coppercable’s physical design contributes to major patch paneland electronic cable management problems. The largerCAT 6A outer diameter impacts conduit size and fill ratioas well as cable management due to the physical size andincreased bend-radius. Copper cable congestion in pathwaysincreases the potential for damage to electronics due to air

cooling damming effects and interference with the abilityof ventilation systems to remove dust and dirt. Opticalcable offers significantly better system density and cablemanagement and minimizes airflow obstructions inthe rack and cabinet for better cooling efficiencies.See Figures 3.2 and 3.3.


Figure 3.2Optical Cable (left) vs. Equivalent Copper Cabling | Photo LAN874

Figure 3.3Copper Cable Management

As network speed grows, optical fiberoffers significant advantages over copper

10 Gbps Example90%

85%

80%

75%

70%

48 96 144

192

240

288

Elec

tric

ityco

stsa

ving

sby

usin

g10

Gop

tical

inst

ead

of10

Gco

pper

(%)

Number of 10G Ports

Figure 3.1Electronics and Cooling Savings

End Equipment Through Optical FiberDistance CapabilitiesSpan length, application and data rate are the determiningfactors in the selection of fiber type and end equipment.All must be considered in order to make the best overallselection. OM3 and OM4 fibers are appropriate for themajority of data center applications, as the associated opto-electronic transmission equipment is usually more economi-cal than that for single-mode systems. Analysis of a specificsystem design will lead to the selection of the most suitablefiber type and end equipment, after which detailed consider-ation of the optical parameters for both fiber and the systemis necessary. The following is a discussion of the nature andmeaning of those optical parameters with which the design-er should be familiar.

TransceiversThe transceiver is an electronic device that receives anelectrical signal, converts it into a light signal and launchesthe signal into a fiber. It also receives the light signal andconverts it into an electrical signal as well. For data rates=>1G, a multimode transceiver uses an 850 nm VCSEL anda single-mode transceiver uses a 1310 nm fabry-perot (FP)or distributed feedback (DFB) laser. Transceivers operatingat 1G and higher data rates migrated from light emittingdiodes (LEDs) to laser sources due to the LED modulationrate limitation and wide spectral width. For systems operat-ing at data rates greater than 622 Mb/s, lasers must be used.VCSEL fabrication and packaging costs are significantly lessthan for a single-mode FP/DFB laser. The relative cost ofan FP/DFB transceiver is typically 2-3 times the cost of an

850 nm transceiver. See Figure 3.4. The 850 nm VCSELtransceiver provides the optimum technical and economicsolution for high bit rate (≥ 1 Gb/s) operation that makesOM3/OM4 the most deployed optical fibers in the datacenter today.

SFP/SFP+ are the dominant transceivers used for data rates1G to 16G (see Figure 3.5). Industry-standard multisourcealliances (MSAs) have defined the transceiver performanceattributes (wavelength, spectral width, Tx power, Rx power,etc.) to insure interoperability and reliability. The SFP/SFP+transceiver performance attributes are incorporated intothe Ethernet and Fibre Channel standards to specify systemrequirements and capabilities. Most transceivers interfacewith LC duplex connectors.

The QSFP transceiver will be used for 40G OM3/OM4Ethernet parallel optics. The optical connector interfacewill be the 12-fiber MPO-style connector. The CXPtransceiver will be used for 100G Ethernet parallel optics.The optical connector interface will be the 24-fiber MPO-style connector. Similar to the SFP/SFP+ transceiver, theQSFP and CXP transceivers performance attributes areincorporated into the 40/100G Ethernet standard to specifysystem requirements and capabilities.


Figure 3.5SFP/SFP+ Transceiver | Drawing ZA-3674

Rela

tive

Cost

3.5850 nm optics1300 nm optics

3.0

2.5

2.0

1.5

1.0

0.5

0.02004 2005 2006 2007 2008 2009

Figure 3.4Relative Cost of Single-Mode vs. Multimode 10G Transceiver

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

OM3/OM4 EMBcFor systems operating at data rates greater than 1 Gb/s,TIA/EIA-455-220 and IEC 60793-1-49 bandwidth testmethods are used to measure the fiber effective modalbandwidth (EMB) that include a series of small spot sizelaunches (approximately 5 µm) indexed across the fiber core.Measurements are made of the output pulse time delay andmode coupling power of the fiber as a function of radialposition. These measurements are referred to as differentialmode delay (DMD) measurements. Data from these meas-urements can be analyzed by two methods to determinewhether the fiber meets the EMB requirement of a specificapplication. The first method for translating DMD meas-urements into an EMB prediction is commonly referred toas the DMD mask approach, where the leading and trailingedges of each pulse are recorded and normalized in powerrelative to each other. This normalization approach reducesthe raw DMD data to focus exclusively on time delay, wherethe overall fiber delay is calculated as the difference betweenthe times for the slowest trailing edge and the fastest leadingedge in units of ps/m. In order for a fiber to be determinedas meeting the required minimum value of 2000 MHz•kmEMB for OM3 at 850 nm, the DMD data must conform toone of six templates or masks and must not show a DMDmeasurement greater than 0.25 ps/m for any of four speci-fied radial offset intervals. In order for a fiber to be deter-mined as meeting the required minimum value of 4700MHz•km EMB for OM4 at 850 nm, the DMD data mustconform to one of three templates or masks and must notshow a DMD measurement greater than 0.11 ps/m for anyof four specified radial offset intervals. It should be notedthat this method provides only a pass/fail estimation againstthe 2000 MHz•km and 4700 MHz•km requirements.

The newer method for predicting EMB from DMD datais called calculated effective modal bandwidth (EMBc). Asmentioned, the DMD measurement characterizes a singlefiber’s modal performance in high detail, including bothmodal time delay and coupling as a function of radialposition. With EMBc, the fiber’s performance is thencharacterized by a series of 10 sources which are chosento span across a range of 10,000 encircled fluxed compliantVCSELs. Conceptually, this is done by weighting theindividual DMD launches to approximate the radial powerintensity distribution of any desired VCSEL. Thoseweightings are then combined with the raw DMD data toconstruct an output pulse for that fiber/laser combination.The resultant output pulse can then be used to calculateEMB in units of MHz•km.

To ensure field performance, EMB is calculated for 10actual laser sources which have been determined to repre-sent the performance extremes of all encircled compliantVCSELs. Of these 10 sources, the one yielding the lowestEMBc value is taken to represent the minimum expectedperformance level of all standards-compliant VCSELs, andthe EMBc value associated with this source is thereforereferred to as the minimum calculated EMB or minEMBc.

The primary advantage of the minEMBc method over theDMD mask method is that the minEMBc method guaran-tees standards-compliant fiber performance under worstcase source/fiber interactions while providing an actualvalue of bandwidth in the scalable units of MHz•km. TheminEMBc value can then be used to calculate bit rates andlink lengths for systems requiring EMB values other than aminimum 2000 MHz•km. Corning Cable Systems recom-mends that multimode fiber intended for current or futureuse at data rates ≥ 1 Gb/s should be specified according tominEMBc values rather than pass/fail performance indicatedby the DMD mask method.


Chapter Four:Standards

General StandardsThere are two types of environments in the data center:local area networks (LANs) and storage area networks(SANs).

A LAN is a network linking multiple devices in a singlegeographical location. Typical LAN speeds are 1 Gb or10 Gb Ethernet.

A SAN is an area in the network linking servers to storageequipment, which introduces the flexibility of networkingto servers and storage. Speeds are typically 2G, 4G, 8G or10G Fibre Channel.

When designing a data center, several factors should betaken into consideration, including standards compliance.TIA-942, Telecommunications Infrastructure Standard forData Center, details several of the factors that should beconsidered when designing a data center. When imple-menting a structured cabling solution, the standard recom-mends a star topology architecture to achieve maximum

network flexibility. TIA-942 outlines additional factorscrucial to data center design, including recognized media,cable types, recommended distances, pathway and spaceconsiderations and redundancy. In addition to standardscompliance, the need for infrastructure flexibility toaccommodate future moves, adds and changes due togrowth, new applications, data rates and technologyadvancements in system equipment must be considered.

Data Center NeedsAs data centers face the continued need to expand andgrow, the fundamental concerns are constant. Datacenter infrastructures must provide reliability, flexibilityand scalability in order to meet the ever-changing datacenter network.

• Reliability: Data center cabling infrastructuresmust provide security and enable 24 x 365 x 7 uptime.Tier 4 data centers have uptime requirements of 99.995percent, less than one-half hour per year.

Chapter Four: Standards | LAN-1160-EN | Page 14

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Designing the Physical Infrastructure 2SECTION


• Flexibility: With the constant in data centers beingchange, the cabling infrastructure must be modularto accommodate changing requirements and easy tomanage and adjust for minimal downtime duringmoves, adds and changes.

• Scalability: Cabling infrastructures must support datacenter growth, both in addition of system electronicsand increasing data rates to accommodate the need formore bandwidth. The infrastructure must be able tosupport existing serial duplex transmission and providea clear migration path to future parallel optic transmis-sion. In general, the infrastructure should be designedto meet the challenges of the data center over a 15- to20-year service life.

TIA-942TIA-942, Telecommunications Infrastructure Standardsfor Data Centers, was released in April 2005. The purposeof this standard is to provide information on the factorsthat should be considered when planning and preparingthe installation of a data center or computer room.TIA-942 combines within a single document all of theinformation specific to data center applications. Thisstandard defines the telecommunications spaces, infra-structure components and requirements for each withinthe data center. Additionally, the standard includes guid-ance as to recommended topologies, cabling distances,building infrastructure requirements, labeling andadministration, and redundancy.

Data Center Spaces and InfrastructureThe main elements of a data center, defined by TIA-942,are the entrance room (ER), main distribution area(MDA), horizontal distribution area (HDA), zonedistribution area (ZDA), equipment distribution area(EDA) and telecommunications room (TR).• Entrance room (ER): The space used for the

interface between data center structured cablingand interbuilding cabling, both access providerand customer-owned. The ER interfaces with thecomputer room through the MDA.

• Main distribution area (MDA): Includes the maincross-connect, which is the central point of distributionfor the data center structured cabling system and mayinclude a horizontal cross-connect when equipmentareas are directly served from the MDA. Every datacenter shall include at least one MDA.

• Horizontal distribution area (HDA):Serves equipment areas.

• Equipment distribution area (EDA): Allocated forend equipment and shall not serve the purposes ofan ER, MDA or HDA.

• Telecommunications room (TR): Supports cabling toareas outside the computer room and shall meet thespecifications of ANSI/TIA-569-B.

The components of the cabling infrastructure, as definedby TIA-942, are as follows:

• Horizontal cabling

• Backbone cabling

• Cross-connect in the ER or MDA

• Main cross-connect in the MDA

• Horizontal cross-connect in the TR, HDA, MDA

• Zone outlet or consolidation point in the ZDA

• Outlet in the EDA

Entrance Room(Carrier Equip and

Demarcation)

ComputerRoom

Main Distribution Area(Routers, Backbone LAN/SAN

Switches, PBX, M13 Muxes)

Telecom Room(Office and OperationsCenter LAN Switches)

Offices, Ops. Center,Support Rooms

Zone Dist Area

Horiz Dist Area(LAN/SAN/KVM

Switches)


Switches)


Switches)


Switches)

Equip Dist Area(Rack/Cabinet)




AccessProviders

AccessProviders

HorizontalCabling

BackboneCabling

Figure 4.1TIA-942 | Drawing ZA-3301

In a data center, including HDAs, the maximum distanceallowed for horizontal cabling is 90 m, independent ofmedia type. With patch cords, the maximum channeldistance allowed is 100 m, assuming 5 m of patch cord ateach end of the channel for connection to end equipment.When a ZDA is used, horizontal cabling distances forcopper may need to be reduced.

Depending on the type and size of the data center, theHDA may be collapsed back to the MDA. This is a typicaldesign for enterprise data centers. In this scenario, thecabling from the MDA to the EDA, with or without aZDA, is considered horizontal cabling. In a collapseddesign, horizontal cabling is limited to 300 m for opticalfiber and 90 m for copper.

TIA-942 defines the maximum distance for backbonecabling as being application and media dependent.

Equip Dist Area(Rack/Cabinet) Zone Dist Area



Switches)

Horizontal Cabling Horizontal

Cabling

Horizontal Cabling

90 m (Horizontal Dist.)100 m (Channel Dist.)

90 m (Horizontal Dist.)100 m (Channel Dist.)

Figure 4.2Horizontal Distribution Area Topology | Drawing ZA-3581

ComputerRoom

Zone Dist Area



Access Providers

HorizontalCabling300 m optical

or 90 m copper




C

Figure 4.3Reduced Data Center Topology | Drawing ZA-3427


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Tier Ratings for Data CentersAdditional considerations when planning a data centerinfrastructure include redundancy and reliability. TIA-942describes redundancy using four tiers to distinguishbetween varying levels of availability of the data centerinfrastructure. The tiers used by this standard correspondto industry tier ratings for data centers, as defined by theUptime Institute. The tiers are defined as Tier I, II, IIIand IV, where a higher tier rating corresponds to increasedavailability. The requirements of the higher-rated tiers areinclusive of the lower level tiers. Tier ratings are specifiedfor various portions of the data center infrastructure,including telecommunications systems architectural andstructural systems, electrical systems and mechanicalsystems. Each system can have a different tier rating,however; the overall data center tier rating is equal tothe lowest of the ratings across the infrastructure.

Tier I Data Center: BasicA data center with a Tier I rating has no redundancy.The data center utilizes single paths and has no redundantcomponents.

From the Uptime InstituteA Tier I data center is susceptible to disruptions from bothplanned and unplanned activity. It has computer powerdistribution and cooling, but it may or may not have araised floor, a UPS, or an engine generator. The criticalload on these systems is up to 100 percent of N. If it doeshave UPS or generators, they are single-module systemsand have many single points of failure. The infrastructureshould be completely shut down on an annual basis toperform preventive maintenance and repair work.Urgent situations may require more frequent shutdowns.Operation errors or spontaneous failures of site infrastruc-ture components will cause a data center disruption.


Redundancy in the Data Center

Offices,Operations Center,

Support Rooms

Telecom Room

Secondary CustomerMaintenance Hole

(Tier 2 and Higher)

Primary Entrance Room(Tier 1 and Higher)

DATA CENTER

TIER

1 TIER3

TIER4TI

ER1

TIER 3 TIER 4

TIER 2 TIER 3

TIER 4

COMPUTERROOM

Secondary Entrance Room(Tier 3 and Higher)

Primary CustomerMaintenance Hole

(Tier 1 and Higher)


Primary Dist Area(Tier 1 and Higher)

Secondary Dist Area(Optional for Tier 4)


Switches) Horiz Dist Area(LAN/SAN/KVM

Switches)


Switches)


Zone Dist Area


Cabling

OptionalCabling

Figure 4.4Tier Ratings for Data Centers | Drawing ZA-3582

• TIA-942 includes four tiers relating tovarious levels of redundancy (Annex G)

• Tier I – No Redundancy– 99.671% available

• Tier II – Redundant component, but 1 path– 99.741% available

• Tier III – Multiple paths, components,but 1 active path– 99.982% available

• Tier IV – Multiple paths, components,all active– 99.995% available– < 1/2 hour downtime/year

Tier II Data Center: Redundant ComponentsA data center with a Tier II rating has redundantcomponents, but utilizes only a single path.

From the Uptime InstituteTier II facilities with redundant components are slightlyless susceptible to disruptions from both planned andunplanned activity than a basic data center. They have araised floor, UPS and engine generators, but their capacitydesign is N+1, which has a single-threaded distributionpath throughout. Critical load is up to 100 percent of N.Maintenance of the critical power path and other parts ofthe site infrastructure will require a processing shutdown.

Tier III Data Center: Concurrently MaintainableA data center with a Tier III rating has multiple paths,but only one path is active.

From the Uptime InstituteTier III level capability allows for any planned site infra-structure activity without disrupting the computer hard-ware operation. Planned activities include preventive andprogrammable maintenance, repair and replacement ofcomponents, addition or removal of capacity components,testing of components and systems and more. For largesites using chilled water, this means two independent sets ofpipes. Sufficient capacity and distribution must be availableto simultaneously carry the load on one path whileperforming maintenance or testing on the other path.Unplanned activities such as errors in operation or sponta-neous failures of facility infrastructure components will stillcause a data center disruption. The critical load on a systemdoes not exceed 90 percent of N. Many Tier III sites aredesigned with planned upgrades to Tier IV when theclient’s business case justifies the cost of additional protec-tion. The acid test for a concurrently maintainable datacenter is the ability to accommodate any planned workactivity without disruption to computer room processing.

Tier IV Data Center: Fault TolerantA data center with a Tier IV rating has multiple activepaths and provides increased fault tolerance.

From the Uptime InstituteTier IV provides site infrastructure capacity and capabilityto permit any planned activity without disruption to thecritical load. Fault-tolerant functionality also provides theability of the site infrastructure to sustain at least oneworst-case unplanned failure or event with no critical loadimpact. This requires simultaneously active distributionpaths, typically in a system-to-system configuration.Electrically, this means two separate UPS systems in whicheach system has N+1 redundancy. The combined criticalload on a system does not exceed 90 percent of N. As aresult of fire and electrical safety codes, there will still bedowntime exposure due to fire alarms or people initiatingan emergency power off (EPO). Tier IV requires all com-puter hardware to have dual power inputs as defined by theInstitute’s Fault-Tolerant Power Compliance SpecificationsVersion 2.0, which can be found at www.uptimeinstitute.org.The acid test for a fault tolerant data center is the abilityto sustain an unplanned failure or operations error withoutdisrupting computer room processing. In considerationof this acid test, compartmentalization requirements mustbe addressed.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Structured CablingTIA-942 provides structured cabling guidance for datacenters. To implement a structured cabling solution, astar topology is recommended. If an unstructured cablingsolution is used (e.g., a point-to-point installation withjumpers), moves, adds and changes (MACs) to the datacenter become difficult. Issues that may arise include thefollowing: manageability, scalability, cooling, density andflexibility. For data centers utilizing access flooring, it isimperative to keep under-floor obstructions like cablingto a minimum so cooling airflow is not impeded.

With a star topology, maximum flexibility in the networkis achieved. TIA-942 states that both horizontal andbackbone cabling shall be installed using a star topology.The cabling infrastructure should be implemented to allowmoves, adds and changes without disturbing the cablingitself. MACs include network reconfiguration, growingand changing user applications and/or protocols.

Figure 5.1Data Center Example | Drawing ZA-3583

EDAServer

Cabinet MDA

SAN

LAN

LAN

EDASAN

SAN Switch

Storage

EDGE Switch

Servers

DistributionSwitch

Router

Figure 5.2Data Center Topology | Drawing ZA-3584

Chapter Five:Designing a Scalable Infrastructure

Chapter Five: Designing a Scalable Infrastructure | LAN-1160-EN | Page 19

Implementation of a star topology with ZDAs allows fora flexible and manageable cabling infrastructure. Cablingcan be consolidated from hundreds of jumpers to just afew low-profile, high-fiber-count trunk cables routed toseveral zone locations. When adding equipment, extendertrunks (usually much lower fiber count than the trunks,i.e., 12 fibers to 48 fibers) can be added incrementally,interconnected at the ZDA (TIA-942 only allows oneZDA in a link; ZDAs cannot be concatenated) and routedto the equipment racks. This can be done easily withoutdisrupting the backbone cabling and without pulling floortiles across the entire data center.

Standards ComplianceWhen designing a data center to meet these needs, bestpractices should be followed. TIA-942 addresses recom-mended design practices for all areas of the data center,including pathways and spaces and the cabling infrastructure.

Design Recommendations Using ZonesZone distribution is not only a design topology recom-mended in TIA-942, but also one incorporated into manydata centers operating today. Consider these steps whenconsidering a zoned architecture:1. Identify zones or zone distribution areas (ZDAs)

throughout the data center.2. Install high-fiber-count cabling from the MDA to the

localized zones or ZDAs.3. Distribute lower-fiber-count cabling from the ZDAs

to the cabinets or components within the zone.

Zone distribution provides many benefits whenincorporated in the data center cabling infrastructure:• Reduces pathway congestion.• Limits data center disruption from the MDA and eases

implementation of MACs.• Enables a modular solution for a “pay-as-you-grow”

approach.

Entrance Room(Carrier Equip and

Demarcation)

ComputerRoom



Telecom Room(Office and OperationsCenter LAN Switches)


Zone Dist Area


Switches)


Switches)


Switches)


Switches)





AccessProviders

AccessProviders

HorizontalCabling

BackboneCabling

Figure 5.3TIA-942 | Drawing ZA-3301

ZDA

ZDA

ZDA

Server CabinetsServer Cabinets



Cabinetsgrouped

into zones

Main DistributionFrame

Main DistributionArea (MDA)

MDF

Zone Distribution Area (ZDA) locatedin the center of each zone

AdditionalCabinetZones

ZA-3585

Figure 5.4Identify Zones or ZDAs | Drawing ZA-3585

ZDA

ZDA

ZDA

Cabinetsgrouped

into zones


MDF

Connectivity is quickly and easily deployed from the ZDAs to the Server Cabinets on anas-needed basis


ZA 3587

Figure 5.6Distribute Lower-Fiber-Count Cabling | Drawing ZA-3587

ZDA

ZDA

ZDA

Cabinetsgrouped

into zones


MDF

Trunk Cabling Star Networkedfrom the MDFs to the ZDAs


ZA-3586

Figure 5.5Install High-Fiber-Count Cabling | Drawing ZA-3586

Zone Distribution in the Data Center

Chapter Five: Designing a Scalable Infrastructure | LAN-1160-EN | Page 20

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

The selection of the fiber count, or number of fibers usedin the cable plant, is an extremely important decision thatimpacts both the current and future system capabilities,as well as the cost of a communications network. Thedevelopment and widespread use of fiber in all aspects ofthe data center network require the designer to plan notonly for the immediate system requirements, but for theevolution of future system demands as well. Since thesefiber systems will provide service for a number of differ-ent applications later, the number of fibers designedinto the network today must be carefully considered.Before fiber counts are determined, the designer needsto analyze the following:

1. Physical infrastructure design for data centers• TIA/EIA 942• Defining MDAs, HDAs and ZDAs

2. Logical topologies for data centers• Common architectures

3.Mapping logical topologies into the physicalinfrastructure• TIA-942 and logical architectures• Choosing the proper TIA-942 architecture

Logical Topologies for Data CenterWhile standards help guide the data center physicalinfrastructure, the data center logical infrastructuredoes not have a standards body helping with design.Logical architectures as shown in Table 6.1 vary basedon customer preference and are also guided by theelectronics manufacturers.

Though a standard does not exist, there are somecommon architecture best practices that can be followed.Most logical architectures can be broken into four layers:

1. Core2. Aggregation3. Access4. Storage

CoreThe core layer provides the high-speed connectivitybetween the data center and the campus network.This is typically the area where multiple ISPs provideconnections to the internet.

AggregationThe aggregation layer provides a point where all serverarea devices can share common applications such as fire-walls, cache engines, load balancers and other value-addedservices. The aggregation layer must be able to supportmultiple 10G and 1 Gig connections to support ahigh-speed switching fabric.

AccessThe access layer provides the connectivity between theaggregation layer shared services and the server farm.Since additional segmentation may be required in theaccess area three different segments are needed:1. Front-end segment – This area contains web servers,

DNS servers, FTP and other business applicationservers.

2. Application segment – Provides the connectionbetween the front-end servers and the back-end servers.

3. Back-end segment – Provides connectivity to thedatabase servers. This segment also provides accessto the storage area network (SAN).

StorageThe storage layer contains the Fibre Channel switches andother storage devices such as magnetic disc media or tape.

Chapter Six:Determining the Fiber Counts

Chapter Six: Determining the Fiber Counts | LAN-1160-EN | Page 21

Layer Logical Architecture

Core

Aggregation

Access

Storage

TABLE 6.1


StorageLayer

Back-EndLayer

AppLayer

Front-EndLayer

Core Layer

AggregationLayer

AccessLayer

StorageLayer

Figure 6.1Logical Architecture | Drawing ZA-3656

TIA-942 Physical Architecture Area Logical Architecture Area

MDA = Main Distribution Area Maps to Core and Aggregation

HDA = Horizontal Distribution Area Maps to Aggregation

ZDA = Zone Distribution AreaMaps to Access and Storage

EDA = Equipment Distribution Area

TABLE 6.2: Mapping Architectures

Mapping Logical Architectures to TIA-942The key for many data center designers is how to translatethe many logical topologies onto a TIA-942 structuredcabling infrastructure. This translation will affect someof the key design elements of a structured cabling solutionsuch as fiber counts, hardware considerations and physicalcable runs. The first step is to translate the TIA-942 areas(MDA, HDA, ZDA, EDA) to the logical architecture areas(core, aggregation, access, storage). Table 6.2 shows acomparison between the two.

The next step is to take an example logical architectureand translate it to a TIA-942 structured cabling solution.In this example, we will use a small data center and mapthe logical architecture shown in Figure 6.1 to the physicalarchitecture of the data center (racks and cabinets) that isshown in Figure 6.2.

The next step is to choose the TIA-942 architecture thatwill best map to the logical architecture shown in Figure6.1. Since this data center is small, a reduced TIA-942architecture will be implemented. In this architecture,an MDA, ZDA and EDA will be implemented.

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS


Figure 6.2Data Center Rack Layout | Drawing ZA-3540

• Core Switching• Aggregation Switching• SAN Switching

Main Distribution Area(MDA)

Server Cabinets

Server Cabinets

Server Cabinets

Storage Cabinets

ZDA

MC

ZDA

ZDA

ZDA

Front-EndLayer Zone EDA EDA EDA EDA EDA EDA EDA EDA

ApplicationLayer Zone EDA EDA EDA EDA EDA EDA EDA EDA

Back-EndLayer Zone EDA EDA EDA EDA EDA EDA EDA EDA

StorageZone EDA EDA EDA EDA EDA EDA EDA EDA

Figure 6.3Data Center Cabled Architecture | Drawing ZA-3541

16x 10GE

32x 10GE

Up To 20 10GE Uplinks Per Switch

2x Switch

2x Blade Server ChassisWith 16 Pass-Through10GE Connections

Figure 6.4Switch Configuration | Drawing ZA-3657

In implementing this structured cabling design, the datacenter will be segmented based on the logical topologyshown in Figure 6.1. The segmentation will be as follows:

1. Collapse the core switching LAN and SAN andaggregation switching in the MDA area.

2. Segment the access layer into three zones (front-end,application and back-end).

3. Segment the storage into a separate zone.

Each zone will use a middle-of-the-rack (MoR)interconnect solution for the cabling and within eachzone, the EDAs will utilize a top-of-the-rack interconnect.The EDAs will serve the electronics in each cabinet andthe ZDAs will serve the EDAs. The ZDAs will homerunback to the MDA where they will terminate in a maincross-connect (MC). This is shown in Figure 6.3.

The next step is to determine the number of fibers thatare needed to implement this structured cabling solution.Two things the designer needs to take into account are:

1. Redundancy requirements for each section or zone2. Networking requirements

Many data centers are set up to have redundant cableroutes to each zone area. An “A” and a “B” route are verycommon in today’s infrastructure design. Redundancy inthe data center will increase the fiber count to each zone.

Networking requirements will also affect the fiber countsin the data center. Many networking configurations willrequire redundant switches in each rack to reduce singlepoints of failure in the data center. Also the numberof upstream ports versus downstream ports (oversubscrip-tion) will affect the fiber count.

As illustrated in the switch configuration shown inFigure 6.4, this configuration calls for two switches on topof the EDA cabinet. Each switch will feed 16 blade serversfor a total of 32 “downstream” ports. The number of“upstream” ports (fiber links back to the MDA) willdepend on how much the network engineers want tooversubscribe the switch. For example, to have a 1:1 over-subscription, you would need 32 upstream ports to matchthe 32 downstream ports. Table 6.3 shows the fiber countsrequired for this configuration.


Oversubscription RatioPer Switch 10G Uplinks Per Switch Fiber Count Per Switch Fibers Per Rack

8:1 4 8 24

4:1 8 16 48

1.6:1 20 40 96

TABLE 6.3: Oversubscription Ratios for 10G

Using Table 6.3 and applying a 1.6:1 oversubscription wouldyield a fiber count configuration shown in Figure 6.4.

In Figure 6.5 each of the nine EDA cabinets require96 fibers to support the oversubscription rate and therequirements for redundancy. Using 144-fiber trunkcables yields three 144-fiber cables to Core A and three144-fiber cables to Core B. The same process would needto be repeated for the other zones in this example.

The Future: 40G/100G SystemsMigrating to the next generation of switches will requirecareful planning for fiber counts. Advanced systems suchas 40G Ethernet and 100G Ethernet will require thousandsof fibers for network connectivity. 40G Ethernet systemswill utilize a 12-fiber MPO-style (MTP®) connector asthe interface into the end electronics. A basic configurationfor a 40G switch may consist of 12 fibers per port and16 ports per card (Figure 6.6).

If the designer replaces the 10G switches with 40Gswitches, the fiber count would increase. Using the samescenario as before (32 servers) and the same oversubscrip-tion ratios as before, the fiber counts per rack increase.Table 6.4 shows the fiber counts based on 40G.

1

2

3

4

5

6

7

8

9

Figure 6.6Switch Configuration | Drawing ZA-3588

EDA EDA EDA EDA EDA EDA EDA EDAZDAFront-EndLayer Zone

Core “A” Core “B”


EDA EDA EDA EDA EDA EDA EDA EDAZDA




Front-EndLayer Zone

ApplicationLayer Zone

Back-EndLayer Zone

StorageZone

96F 96F 96F 96F 96F 96F 96F 96F

3 x 144F 3 x 144F



Figure 6.5Fiber Count Configuration | Drawing ZA-3658

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS



8:1 4 48 72

4:1 8 96 144

1.6:1 20 240 288


Using Table 6.4 and applying a 1.6:1 oversubscription wouldyield the fiber count configuration shown in Figure 6.7.

In this example each of the nine EDA cabinets require288 fibers to support the oversubscription rate of 1.6:1and the requirements for redundancy. Using 144-fibertrunk cables yields nine 144-fiber cables to Core A andnine 144-fiber cables to Core B.

100G Ethernet systems will utilize a 24-fiber MTP®

Connector as the interface into the end electronics.A basic configuration for a 100G switch may consistof 24 fibers per port and 16 ports per card.

If the designer replaces the 10G switches with 100Gswitches, the fiber count would increase. Using the sameoversubscription ratios as before, the fiber counts per rackincrease. Table 6.5 shows the fiber counts based on 100G.

Using Table 6.5 and applying a 1.6:1 oversubscriptionwould yield a fiber count configuration shown inFigure 6.8.

In this example, each of the nine EDA cabinets require576 fibers to support the oversubscription rate of 1.6:1and the requirements for redundancy. Using 144-fibertrunk cables yields 18 144-fiber cables to Core Aand 18 144-fiber cables to Core B.


8:1 4 96 144

4:1 8 192 288

1.6:1 20 480 576






Front-EndLayer Zone


Back-EndLayer Zone

StorageZone

288F 288F 288F 288F 288F 288F 288F 288F

9 x 144F 9 x 144F



EDA EDA EDA EDA EDA EDA EDA EDAZDAFront-End







Front-EndLayer Zone


Back-EndLayer Zone

StorageZone

576F 576F 576F 576F 576F 576F 576F 576F

18 x 144F 18 x 144F




In commercial building installations, an optical fibercabling link is typically assembled in the field at the jobsite. The cable is pulled in from a reel of bulk cable, cut tolength, attached to the patch panel housing and terminatedwith field-installable connectors on each end. The termi-nated ends are then loaded into adapters in rack- or wall-mountable housings. Finally, the complete link is tested forcontinuity and attenuation.

Alternatives to the traditional implementation methodare factory-terminated and pre-assembled solutions.The time-consuming steps of installation, such as cablesheath removal, cable furcation, connector installationand hardware assembly, can be completed in the factory.The complete package is shipped to the job site readyfor quick and easy installation.

Preterminated cables and hardware are ideal for use indata centers to enable utilization of high-density fiber ports.In these applications and others, the benefit of a pre-assem-bled solution saves installation time, reduces system down-time and provides a more flexible and scalable solutionwith high-quality factory terminations than traditional fieldinstallation methods.

With little planning prior to ordering, preterminatedsolutions offer several advantages over the traditionalinstallation such as:

• An optical fiber link can be quickly and easily installed.This can be most advantageous for projects where systemdowntime must be minimized or where disruption of thefloor space cannot be tolerated. Hence, a pre-assembledsolution can be useful to have on hand for emergencyrepairs or for the re-cabling of a facility that must remainoccupied and functional.

• A pre-assembled solution can be useful where costcontrol of a project is most important. The completionof many of the labor assembly steps at the factory cansignificantly reduce the variability of installation costin the field.

• A pre-assembled solution can increase the versatility andproductivity of the installation crew with fewer demandson specialized tooling and installation skills.

• An optical fiber link component can be completely assem-bled and tested prior to leaving the factory. Most of theproblems associated with the traditional field installationoccur with the field connectorization and correct loadingof the connectors into the hardware. These problems aregreatly reduced with factory-terminated connectors.

The design and product selection process remains the samewith selection and specification of fiber type, fiber count,cable type, connector type and hardware type appropriatefor the environment.

The following additional steps are required:

1. Predetermine the installed link length.

2. Ensure that a connectorized cable end can be pulledthrough the path (i.e., conduit) or space for the cableroute. This must take into account the pulling grip size.

Corning Cable Systems preterminated solutions arewell-suited for the data center for the following reasons:

• Denser optical networking solutions, free raised floorand rack space

• Significantly faster installation times

• Modular design for faster moves, adds and changes

• Factory terminations

1. Consistent results from an ISO 9001 and TLQ9000certified factory

2. 100 percent factory tested

• Defined accountability, built-in compatibility

• Elimination of variability in material and installation costs

Chapter Seven: Choosing the Infrastructure Components | LAN-1160-EN | Page 26

Chapter Seven:Choosing the Infrastructure Components

Preterminated Solutions

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Deploying the Physical Infrastructure 3SECTION


Standard-Density SolutionsWhen standard density requirements apply, Corning CableSystems recommends Plug & Play™ Universal Systems forthe effective implementation of a factory-pre-assembledsolution. Plug & Play Universal Systems integrate many ofCorning Cable Systems high-quality components, cables,connectors and hardware into a variety of preterminated,pre-assembled and factory-tested solutions while seamlesslymanaging polarity.

MTP® Multi-Fiber ConnectorPlug & Play Universal Systems utilize the MTP®

Connector. The MTP Connector is a multi-fiber array-styleconnector that can accommodate up to 12 fibers in roughlythe same size and footprint as an SC connector. It has asingle high-density footprint of 25 x 10 mm and featuressimple push-on/pull-off mating. A general industry termfor this style of connector is MPO. This connector, whichis used in both multimode and single-mode applications,maximizes valuable panel and hardware space, ensuringhigh density. MTP Connectors are manufactured witheither alignment pins or with alignment holes to ensureproper alignment of the fibers. A connector with alignmentpins always mates with a connector with alignment holes.(Figure 7.1). The MTP Connector offers:

• Up to 54 percent reduction in pathway congestion

• Modularity and scalability with a fiber count that mapsto current and future line-card port counts

• Universal wiring and superior loss performance formigration to higher data rates

Plug & Play Systems ComponentsCable Trunk AssembliesA traditional Plug & Play Universal Systems trunk consistsof an optical cable with each end factory-terminated withMTP Connectors and a pulling grip on one or both ends.Trunks are available in a variety of fiber types and typicallycarry a plenum rating unless otherwise specified. Whenordering Plug & Play Universal Systems trunks, the MTPConnectors on both ends will have pin alignment holes.This ensures that it will integrate with the remaining partsof the system that have pins. It should be noted that MTPConnector panels have neither pins nor alignment holes, asthey are connection points for various components of thePlug & Play Universal Systems.

To successfully deploy a cable that is preterminated on bothends, it is necessary to accurately predetermine the installedlink length. This can be relatively straightforward if well-defined pathways and spaces exist for the cable route, whichis usually true for the data center environment.

If the route is less defined, preterminated cables can stillbe utilized by specifying the trunk cables be longer thanthe known length and planning for the storage of excesscable loops.

PUSH

CONEC

Pul l TO

REMOVE

Figure 7.1MTP Multi-fiber Connector | Drawing ZA-1572


Pulling GripsThe factory-terminated connectors on the Plug & Play™

Universal Systems trunk cables are protected with aprotective pulling grip (Figure 7.2). This grip is designedto be installed by being hand-pulled through a duct, under araised floor and through a riser shaft or through a droppedceiling. The pulling grip is rated to a 100 lb tensile load.

In a data center environment, cable will be pulled through avariety of pathways such as overhead trays, raised floor, conduitand fiber troughs. The size of the pulling grip, the number ofturns in a pull and existing cable fill will dictate the maximumsize of the cable as well as the overall cable pull strategy.

The pulling grip comes in various outer diameters andwill accommodate single-fiber connectors and duplex small-form-factor connectors, such as LC and MTP® Connectors.When the grip is open, the connectors are in a protectivemesh. After the grip is populated with connectors, it is closedand encased inside the black mesh netting that facilitatespulling the cable. The grip has a tensile strength of 100 lbthat provides more than sufficient pull strength for mosthand-pull installation requirements.

Mounting features in the furcation plugs integrate the trunkcable mounting into Corning Cable Systems hardware. Thedesign allows you to attach assemblies quickly into equipmentracks or cabinets with optional mounting brackets. It iscritical to reiterate that a designer must take into account thesize of the pulling grip when determining conduit and otherconstricting pathways.

Extender TrunksExtender trunks are typically used to distribute portions, orall, of the fibers in a Plug & Play Universal Systems trunk toother areas of the infrastructure, such as in a zone distributionarea (ZDA). The extender trunk interconnects to a standardPlug & Play Universal Systems trunk via an MTP Connectoradapter panel. An extender trunk physically differs from astandard Plug & Play Universal Systems trunk in that it hasan MTP Connector with pins on one end and an MTPConnector with alignment holes on the other. The end withpins mates to the standard Plug & Play Universal Systemstrunk, which has alignment holes to accept the alignmentpins. The extender trunk extends the reach of the network byacting like an extension cord. For example, the pinned end ofthe extender trunk plugs into the alignment holes of a maintrunk, while the alignment holes on the extender trunk’sother connector plugs into a Plug & Play Universal Systemsmodule or harness that is pinned.

Hybrid Connector Trunksand Hybrid Extender TrunksPlug & Play Universal Systems hybrid connector trunksare terminated with MTP Connectors on one end of thetrunk and discrete LC or SC connectors on the other endfor applications requiring one end of the trunk to connectdirectly into system equipment or patch panels. Storagedevices, for example, are often stand-alone units that donot have rack space for mounting patch panels. Additionally,if floor space is not available to add a ZDA, the hybridconnector trunk or extender trunk would be a viable option.Hybrid trunks would plug into a module on the MTPend and into electronics panel on the discrete fiber end.The hybrid extender trunk would plug into a trunk on theMTP end and into electronics panel on discrete connectorend (Figure 7.3).

Figure 7.3Plug & Play Hybrid Trunk, 144 fiber | Photo LAN801

Figure 7.2Plug & Play Universal Systems Pulling Grip withIntegrated Mounting Hardware | Photo LAN654

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

ModulesPlug & Play™ Universal Systems modules transition from theMTP® Connector on the trunk cable to the discrete connec-tors used in electronics. For example, if an LC duplex con-nector is used on the edge switches, an MTP-to-LC duplexmodule requires only one panel space in the hardware for upto 24 fibers, utilizing that valuable real estate wisely by usingLC duplex connectors to double the capacity of the boxinstantly (Figure 7.4).

Note: Polarity is discussed in detail in Chapter Eleven. Corning Cable SystemsApplication Engineering Note 69, “Plug & Play Universal Systems” alsoaddresses polarity.

Integrated Trunk ModuleThe Integrated Trunk Module (ITM) is an innovativedesign that incorporates the benefits of a module withdiscrete connectors with the advantage of an integratedtrunk cable. Figure 7.5 shows the module in detail. TheIntegrated Trunk Module is a preterminated 12-fiber MTPConnector trunk assembly integrated into a Plug & PlayUniversal Systems module. The trunk cable stored withinthe module is easily deployed to an exact length, so precisepre-planning of cable length is not required. It is ideal forthe zone distribution area in a large data center as it pro-vides a quick and convenient method for deploying and/orre-deploying optical connectivity. In a small data center, itis perfect for connecting the main distribution area to thesystem equipment cabinets.

HarnessesLike modules, harnesses allow the user to maintainmodularity with a system solution implemented byproviding a transition from the MTP Connectors used onthe trunk or extender trunk cables to single- or dual-fiberconnectors. A harness is a cable assembly with a multi-fiberpinned MTP Connector on one end and simplex or duplexconnectors at the ends of up-jacketed legs.

Rather than terminate the end of a trunk into a module,the trunk is interconnected to a harness through an MTPConnector adapter panel. The individual connectors atthe opposite end of the harness assembly can be pluggeddirectly into equipment or into patch panels, with patchcords used to provide connectivity to equipment (Figure 7.6).


Figure 7.4Plug & Play Universal Systems Module | Photo LAN1797

Figure 7.5Integrated Trunk Module | Photo LAN1680

Figure 7.6Harness | Photo LAN1371

Patch CordsPatch cords can be purchased with various options for thefiber and connectors used. In a data center installation thatis implemented with structured cabling in mind, patch cordsor jumpers should only be used to provide connectivitybetween end equipment and trunk cables, or cross-connec-tions between trunk cables. Long-length patch cords shouldnot be used as the primary method to install cabling betweenpieces of equipment located in various areas of the datacenter. This is because the type of cable used in patchcords is a lighter construction than distribution cablesand cannot stand up to the heavy usage environment ofunder-floor raceway or overhead ladder racks.

HardwareIn addition to the modular cabling components of thePlug & Play™ Universal Systems solution, hardware choicesfor the data center must be considered. Typically the maindistribution area is very dense and requires a higher-densityrack-mount solution. When implementing structured cablingwithin a zone distribution area, a low-profile solution isdesirable within a rack. Other zone locations include abovethe rack in the cable tray or below the rack underneatha raised floor.

For a cabinet or rack solution, Corning Cable Systemsrecommends the Pretium® Connector Housing (PCH)(Figures 7.7, 7.8 and 7.9) or the dense 1U 96-fiber shelf(Figure 7.10). The PCH is available in rack heights of 1U,2U and 4U. The PCH includes an additional four inches ofdepth for increased slack storage space inside the housings,as well as a 1U integrated horizontal jumper managementfor the PCH-04U. The 96-fiber shelf comes configuredspecifically for LC duplex connections.


Figure 7.8PCH-02U | Photo LAN1008

Figure 7.7PCH-M3-01U | Photo LAN994

Figure 7.9PCH-04U | Photo LAN1399

Figure 7.101U 96-Fiber Shelf | Photo LAN1248

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS


For an overhead or sub-floor zone distribution solution,Corning Cable Systems offers the Fiber Zone Box (FZB-04U) which provides module capacity and/or up to 12-panelcapacity. The FZB-04U fits through a standard 2 x 2 ft flooror ceiling tile (for overhead installations). In either solution,the Fiber Zone Box provides a space for the interconnectionof cabling, via modules or panels. The FZB-04U accepts upto 4U of 19-in rack-mountable equipment inside it, allowingfor combining copper patch panels in the same housing asfiber patch panels (Figure 7.11).

Another type of hardware to use when space is a concern isa low-profile bracket that can be integrated into equipmentcabinets that accept modules or panels, such as theRBC-02P or CPP-01U-PNL (Figures 7.12 and 7.13).

Figure 7.12RBC-02PwithModule | Photo LAN1208

Figure 7.13CPP-01U-PNL Low-Profile Bracket | Photo LAN1361

Figure 7.11Fiber Zone Box with Blank Panels | Photo LAN589

Figure 7.14Pretium EDGE Solutions | Photo LAN1740

Figure 7.16B144-Fiber Trunk Cable | Photo LAN1568

Figure 7.17Trunk Furcation Comparison | Photo LAN1551

Standard Plug & Play Systems Furcation

Figure 7.15Pretium OM3 Jumper | Photo LAN1528

Figure 7.16ATrunk Pulling Grip | Photo LAN1569

Pretium EDGE Solutions Furcation

High-Density SolutionsWhen high density is a requirement, Corning Cable Systems recommendsPretium EDGE™ Solutions. Pretium EDGE Solutions, a complement tothe Plug & Play™ Universal Systems product family, provides increasedsystem density when compared to traditional preterminated systems andoffers the highest port density currently in the market. Custom engineeredcomponents enable simple integration into common storage area network(SAN) directors, while the preterminated components allow for reducedinstallation time, as well as faster moves, adds and changes (MACs).

Factory-terminated solutions including both Plug & Play Systems andPretium EDGE Solutions provide improved system performance, ensurecomponent compatibility and yield consistent quality. Pretium EDGESolutions consist of optical trunks, extender trunks, modules, harnesses,housings and jumpers. Enabled by reduced cable diameters and Corning®

ClearCurve® multimode optical fiber (Figure 7.15), the trunks and extendertrunks have an innovative pulling grip to increase the speed of deploymentwhile offering superior protection of the assembly (Figure 7.16A and7.16B). Trunk furcation is smaller than its predecessor (Figure 7.17) andeasily integrates inside the hardware via a cradle to create a rapid one-handstrain-relief system.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

The universally wired modular system components enablefast and simple networking moves, adds and changeswith none of the polarity concerns associated with specialpolarity-compensating components.

Pretium EDGE™ Solutions ComponentsTrunk CablesPretium EDGE Solutions trunk cables (Figure 7.18A) utilizethe MTP® Connector and support 12 to 144 fibers. PretiumEDGE Solutions result in up to 65 percent space savings.

Five to six times the fiber tray capacity can be achieved overtraditional bulkier cabling solutions while minimizing cabletray weight and cooling air impediment. The trunk cablescontain Corning® ClearCurve® multimode optical fiber,which enables a bend-radius of five times the cable outsidediameter and allows for smaller slack storage coils or loops.The cables feature a 2.9 mm round furcation leg whichprovides easy routing and improved slack storage.

Pretium EDGE Solutions trunks feature a furcation plugdesign that offers stress-free strain-relief of the cable, anda small-profile furcation plug allows installation of all fibercounts inside a 1U housing.

All trunks are shipped with appropriate strain-relief bracketsfor integrating into PretiumEDGE Solutions housings.The trunk furcation features a transition boot for smoothtransition out of rack/floor hardware. The pulling grip allowsthe trunk to be easily installed around the corners of tray andladder racks, while its robust design allows the trunk to bepulled through conduit using up to 100 lb of pulling tensionwhile providing complete protection for the connectors.All trunks are packaged on a plastic corrugated reel foreasy installation. This reel can be easily broken down forcost-effective disposal (Figure 7.18B).

Pretium EDGE Solutions trunks have been tested and meetthe skew criteria to ensure the system is 100G Ready.


Figure 7.18APretium EDGE Solution Trunk, 12-Fiber | Photo LAN1548

Figure 7.18BPretium EDGE Solutions Trunk Reel | Photo LAN1567

Extender TrunksPretium EDGE™ Solutions extender trunks are used toextend the Pretium EDGE Solutions trunk cables from azone consolidation area. For example, a high-fiber-counttrunk can be deployed from a main distribution area(MDA) to a zone distribution area (ZDA). Smaller-fiber-count extender trunks can then be utilized to distributefiber into multiple cabinets within a row. Network design-ers can build the backbone cable (trunks) to full capacityby utilizing this design. As equipment is added, extendertrunks are deployed. This ensures that the core data centerwill experience very little of the disruption normallyassociated with point-to-point design philosophy.

Extender trunks are manufactured with pinned MTP®

Connectors on one end of the cable trunk and non-pinnedMTP Connectors on the other end. The pinned MTPConnectors mate with the non-pinned connectors of thePretium EDGE Solutions trunk and the non-pinned MTPConnectors are plugged into the Pretium EDGE Solutionsmodule or harness.

ModulesPretium EDGE Solutions modules are used to breakout the 12-fiber MTP Connectors terminated on trunkcables into LC connectors to facilitate patching intosystem equipment ports, patch panels or work area outlets.The 12-fiber module features LC port adapters acrossthe front and an MTP Connector adapter in the back(Figure 7.19). A factory-terminated and -tested opticalfiber assembly inside the module connects the frontadapters to the back MTP Connector adapter.

PretiumEDGE Solutions modules are housed at the frontof the housing and may be installed or removed fromeither the front or rear. This results in faster installationby allowing a technician to strain-relieve the trunk, routetrunk legs, plug the MTP Connector into the moduleand install the module into the housing all from one sideof the cabinet row.

Figure 7.20Trunk Installation into Pretium EDGE Solutions Modules| Photo LAN1751

Figure 7.19Pretium EDGE Solutions Module | Photo LAN1542


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Figure 7.21VFL-Compatible Shutter | Photo LAN1545

Figure 7.22Module in Easy-Open Plastic Packaging | Photo LAN1543

The module features an LC shuttered adapter, so there is noneed for dust caps, which can be lost. Unlike other shutteredadapters on the market, this adapter is VFL compatible.The innovative design diffuses the red VFL light, allowingeasy visualization of port identification without needing tomanually open each shutter (Figure 7.21).

The diffusion property of the door material also provideslaser safety for the technician. This revolutionary inwardopening design also allows for a single-hand LC duplexoperation, while its concave shutter door ensures there isno contact with the connector end face during installation.Pretium EDGE™ Solutions modules are packaged in aneasy-open plastic container to facilitate bulk packaging forreduced waste during data center installation (Figure 7.22).

The use of Pretium EDGE Solutions modules in the datacenter offers the advantage of greater manageability andflexibility with a modular infrastructure. As future connec-tivity requirements change, modules can be easily exchangedto meet those needs, while leaving the existing trunk cableinfrastructure in place. The MTP® Connector backplaneallows for future upgradeability to parallel optics.


HarnessesPretiumEDGE™ Solutions harnesses are used to break outthe 12-fiber MTP® Connectors terminated on trunk cablesinto LC connectors (Figure 7.23). With a pinned MTPConnector on one end that connects to a Pretium EDGESolutions trunk, the other end is equipped with LC-styleuniboot connectors that plug into electronic ports. The useof harnesses provides a solution that occupies less space thantraditional jumpers, as the cable end of the harness is muchsmaller than six equivalent patch cords. This reduced cablingbulk improves airflow for increased cooling and facilitateseasier moves, adds and changes (MACs).

Utilizing 12-fiber MTP and LC uniboot connectortechnology, the Pretium EDGE Solutions harnesses reducecable congestion in front of the SAN director for easyMACs. The harnesses feature a custom-engineered taperto match the port pitch in the electronics to provideseamless integration between the cabling infrastructureand electronics. Furcation plugs can be snapped togetherto maximize harness organization in front of the electronicports, and they contain an integrated Velcro strap hoop toimprove the cabling aesthetics of the densest SAN directors(Figure 7.24).

Pretium EDGE Solutions harnesses are available in twolengths. Short harnesses allow for minimal cable slack whenplacing the electronics adjacent to the housing containingthe MTP Connector interconnect panels. Longer harnesslegs allow flexibility to mount the hardware and electronicsanywhere within the same cabinet, and the MTP Connectorleg slack can be stored in the vertical manager.

MTP Connector PanelsPretium EDGE Solutions MTP Connector panels areused to provide a convenient interconnect point betweenthe trunk cables and harness or trunks and extender trunks.The panels are available with two or four MTP adapters,providing interconnect for 24 or 48 fibers (Figure 7.25 and7.26). Pretium EDGE Solutions MTP Connector panelsare housed within the front of the 1U or 4U housings. Thepanels may be installed or removed from either the frontor rear direction. The MTP Connector backplane allowsfor future upgradeability to parallel optics. Pretium EDGESolutions MTP Connector panels are packaged in a recy-clable easy-open plastic container to facilitate bulk packag-ing for reduced waste during data center installation.

Figure 7.26Pretium EDGE Solutions MTP Connector Panels, 48-Fiber| Photo LAN1795

Figure 7.25Pretium EDGE Solutions MTP Connector Panels, 24-Fiber| Photo LAN1546

Figure 7.24Pretium EDGE Solutions Harnesses Installed in Electronics| Photo LAN1536

Figure 7.23Pretium EDGE Solutions Harnesses | Photo LAN1554


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Figure 7.27Pretium EDGE Solutions 4U Housing | Photo LAN1564

Figure 7.28Pretium EDGE Solutions Housing, Rear View | Photo LAN1746

Figure 7.29Cable Entry into Pretium EDGE Solutions Housing | Photo LAN1747

HousingsPretium EDGE™ Solutions housings are available in both1U and 4U sizes and mount in 19-in racks or cabinets.Combined with Pretium EDGE Solutions trunks, modulesand jumpers, they provide industry-leading high-densityconnectivity, with a port density of 576 fibers within a single4U housing (Figure 7.27). The housings are highly config-urable to meet the dynamic connectivity environments ofboth the main distribution area (MDA) and equipmentdistribution area (EDA) locations in the data center.

In the MDA, the Pretium EDGE Solutions housing providesa cross-connect for first level backbone cables, entrancecables and equipment cables. In the EDA, the housing is aninterconnect to system equipment (such as SAN switches,servers and IP switches). This allows for the effortless addi-tion of groups of switches, storage devices or servers with itsmodular design. The housing also accommodates the mixingand matching of multiple Pretium EDGE Solutions modulesand panels within one chassis.

The 4U housing contains 12 individually sliding trays andthe 1U housing contains two trays. Each tray can accommo-date as many as four 12-fiber modules, resulting in improvedfinger access to connectors and allowing for individual accesswithout compromising the optical connectivity of otherports. This feature leads to fast and simple moves, addsand changes of port configurations. Each sliding tray canaccommodate any of the following port configurations:

• Four Pretium EDGE Solutions 12-fiber modules that canbe installed from the front or the rear of the housing.

• Four Pretium EDGE Solutions MTP® adapter panels.


Front jumper management guides on each tray allow forjumpers to be routed to the left or right (Figures 7.30 and7.31). External jumper routing guides facilitate proper slackmanagement to ensure drawer movement. These guidesare designed at the pivot point between the extreme drawerpositions. Industry-leading port labeling is available on therear side of the housing’s front door, and each housingincludes mounting plug receptacle areas at the rear, whichenable fast and easy installation and strain-relief of PretiumEDGE™ Solutions trunk cables. Brush cable entry makestrunk entry quick and easy.

JumpersCorning Cable Systems offers the most complete line ofconnectors and factory-terminated cables, including jumpersthat meet or exceed all industry standards for reflectance andinsertion loss. Corning Cable Systems’ advanced, state-of-the-art manufacturing process ensures unsurpassed jumperperformance. Fibers and ferrules are thoroughly screened atthe beginning of every process, assembled and polished ina carefully monitored and controlled process, and tested toensure the same outstanding quality in every connector.

Pretium EDGE Solutions jumpers (Figure 7.32) are highlyflexible, easily routed assemblies that use a small-diameter2-fiber interconnect cable to improve the managementof high-density applications. Containing bend-insensitivemultimode and single-mode fiber, the jumpers aredesigned to withstand tight bends and challenging cableroutes. Pretium EDGE Solutions jumpers improve bendtolerance without sacrificing critical bandwidth capabilityor requiring any adjustments to standard field installationor maintenance procedures.

Figure 7.32Pretium EDGE Solutions Jumper | Photo LAN1547

Figure 7.31Pretium EDGE Solutions External Jumpers | Photo LAN1732

Figure 7.30Pretium EDGE Solutions External Jumpers | Photo LAN1783


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

This section will discuss how to choose the properspecifications needed to create a good request for proposal(RFP). The RFP process is an important step in insuringthat the data center designer procures the proper product.The steps needed to implement an RFP are:

1. Pre-workFigure out what you really need, what you want, andwhat is possible for the data center design.

2. Distinguish between needs and wantsUse proper wording to make sure you can separate“needs and wants.”

3. Decide what the winner will look likeEach RFP response will be different. Each companythat responds will have different strengths.

4. Organize the documentMake sure the document has a logical flow and thatthe points are clear.

5. IntroductionExplain to potential bidders why you are publishingthe RFP.

6. RequirementsThis section is the most important and it usually takesthe most time. Make good use of generic specifications(see examples in this chapter) to help you write correctrequirements.

7. Selection criteriaIn this section you tell the bidders how the winningbidder will be selected.

8. TimelinesThis section tells companies who want to bid on yourRFP how quickly they must act and how long theprocess may take. This is also where you tell thebidders how long the evaluation process will take.

9. ProcessIn this section you explain how the process will work-from sending out the RFP to awarding the contractand starting the work.

10.Decide how to send out the RFPMost RFPs are mailed, but you can send the RFPby e-mail or post it on your company website.

11.Decide to whom to send the RFPYour company’s list of acceptable vendors.

12. Send the RFP

Generic Specification Example:Pretium EDGE™ SolutionsCorning Cable Systems reserves the right to update thesespecifications without prior notification.

Pretium EDGE Solutions: GeneralPretium EDGE Solutions include factory-terminatedsystem components that can be quickly mated to forman end-to-end optical link between patching locationsand/or equipment ports. Pretium EDGE Solutions arehigh-density system solutions with reduced installation time.

• Pretium EDGE Solutions are modular solutions withfiber trunks terminated with 12-fiber MTP® ArrayConnectors that mate at each end to a transition harnessor transition module. Harnesses are cable assembliesthat transition from a 12-fiber MTP Array Connectorto single-fiber connectors. Modules have an identicalconfiguration and they are protected in a modular case.Modular system solutions offer a greater degree offlexibility in managing equipment moves, adds, orchanges. An example of this type of system is givenin Figure 8.1.

To maintain proper system polarity, components shall bespecified to comply with universal wiring as described inChapter 11 for new builds.

Insertion loss specifications of individual componentsrepresent the expected performance when mated to othersystem components of like specification.

Chapter Eight:Writing a Data Center Request for Proposal

Chapter Eight: Writing a Data Center Request for Proposal | LAN-1160-EN | Page 39

Pretium EDGEModule

MTP-terminatedTrunk Cable

OptionalMTP-terminatedExtender Trunk

MTP to LC DuplexHarness

Transceiver

Tx/Rx

Rx/Tx

Transceiver

LC DuplexPatch Cord

Figure 8.1Modular Pretium EDGE Solutions Connected to TransceiverPorts with Jumpers and Harnesses | Drawing ZA-3667

Trunk Specifications and OptionsTrunk Function and Construction• The operational temperature range for trunks shall

be -10° to +60°C.• Trunks shall be all-dielectric construction.• Trunks shall be constructed with MTP® Connectors

at both ends.• Trunk fiber count shall be specified as 12, 24, 36, 48,

72, 96, or 144.• Trunks shall be furcated (subdivided) into 12-fiber legs

(subunits). Standard leg length shall be 33 in +3/-0 in.• Trunk length shall be specified as the distance between

furcation points at each end of the cable and shall notbe inclusive of the length of the legs at each end.

• Trunk furcation plugs shall consist of a molded outershell filled with an epoxy encapsulant.

• The furcation plugs shall be square in order to facilitateplug rotation in 90 degree increments. This featureallows mounting the trunk into the hardware in anyorientation and avoids standing torsional forces appliedto the cable.

• There shall be two plug sizes depending on the fibercount trunk. Trunks with 12 to 36 fibers shall beconstructed with a size 1 plug. The size 1 plugdimensions shall be 14.7 mm x 14.7 mm x 108.6 mm.The plug shall have a saddled area with dimensions of11.5 mm x 11.5 mm x 46 mm in order to accommodatea field-installable snap-on device to secure the plug intothe hardware. Trunks with 48 to 144 fibers shall beconstructed with a size 2 plug. The size 2 plugdimension shall be 20 mm x 20 mm x 108.6 mm.The plug shall have a saddled area with dimensions of16.8 mm x 16.8 mm x 46.6 mm in order to accommodatea field-installable snap-on device to secure the plug intothe hardware.

• The trunk shall incorporate a flexible boot at the back ofthe epoxy plug, in order to provide a uniformly smoothtransition between the plug and the trunk cable.

• A tool-less snap-on device shall be used to secure thetrunk into the hardware. There shall be three types ofsnap-on devices depending on fiber trunk count andapplication. For low-fiber-count trunks (12 through36 fibers), single and double stack snap-on devices shallbe offered. The double stack snap-on devices allow youto secure twice the trunk density within the hardware.Single stack snap-on devices shall be available for trunkswith fiber counts greater than 36.

• Trunk furcation plugs shall provide a mounting pointfor a protective pulling grip and shall be capable ofsustaining the rated tensile load of 100 lbs.

• Trunk furcation plugs shall incorporate mechanicallydesigned features that allow securing the trunks insideor outside a connector housing.

• The trunk components shall be RoHS compliant.• Trunk cables shall be manufactured with ultra-bendable

fiber and meet the fiber performance mentioned inTable 8.2.

• The trunk cable shall have a minimum bend-radius offive times the cable’s outside diameter.

• The trunk cable shall meet the application requirementsof the National Electric Code® (NEC® Article 770)OFNP and FT-6 listed for plenum.

• The trunk cable shall meet the outer diameters specifiedin Table 8.1.

Fiber Trunk Count Trunk Cable OD (mm)12 4.824 6.436 7.048 7.672 9.896 10.4144 11.5

TABLE 8.1: Trunk Cable Outer Diameter

TABLE 8.2: Available Fiber Types, Optical Specifications, Jacket Colors for Trunks

Notes:1) As predicted by RML BW, per TIA/EIA 455-204 and IEC 60793-1-41, for intermediate performance laser-based systems (up to 1 Gb/s).2) As predicted by minEMBc, per TIA/EIA 455-220 and IEC 60793-1-49 for high-performance laser-based systems (up to 10 Gb/s).

Multimode Single-Mode

Priority Pretium® 300Ultra-BendableOptimized 50µm (850/1300nm)

Pretium 500Ultra-BendableOptimized 50µm (850/1300nm)

Bend-Improved Single-Mode (1310/1550 nm)

Fiber Attenuation,max (dB/km) 3.0/1.0 3.0/1.0 0.4/0.3

Minimum Over Filled Launch(OFL) Bandwidth (MHz•km) 1500/500 3500/500 -/-

Minimum Effective ModalBandwidth (EMB) (MHz•km) 2000/- 4700/- -/-

Jacket Color Aqua Aqua Yellow


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

• The trunk legs shall be round and have a 2.9 mm outerdiameter with no preferential bend for easy routing.

• Trunks shall meet the connector performancespecifications of TIA/EIA-568-C.3, Optical FiberCabling Components Standard, (normative) Annex A.

Trunk Fiber Types, Optical Specifications,and Jacket Color• Available fiber types and their optical performance

specifications shall be as indicated in Table 8.2.• Trunk jacket color shall be as indicated in Table 8.2.

Trunk Connectivity• Where modular trunks are specified, connectors shall be

MTP having 12 fibers per ferrule.• MTP-terminated primary trunks shall have non-pinned

MTP Connectors on both ends.• MTP-terminated extender trunks shall have pinned

MTP Connectors on the end to be interconnected with aprimary trunk and non-pinned MTP Connectors on theother end.

Trunk Protective Pulling Grips and Covers• Both ends of a trunk shall have a protective packaging

over the furcation plug, legs, and connectors. Customermay specify a protective pulling grip on one end, bothends, or neither end.

• Pulling grips shall be fastened to the epoxy furcationplug in a manner that isolates the cable assemblycomponents (connectors and legs) from tension, torsion,crush, and bending loads encountered when followingrecommended installation practices.

• Pulling grips shall withstand a maximum pulling forceof 100 lbs.

• Trunk pulling grip diameter and minimum allowablebend-radius shall be as indicated in Table 8.3.

• The pulling grip shall be a three components design.The components include a zipper bag, a corrugatedtube and two coupling shelves that allow quick and easyremoval of the pulling grip.

Trunk Packaging• The trunk shall be packaged in a corrugated plastic reel.

The trunk shall be secured to the reel with shrink wrap.• The plastic reel shall be constructed with 100%

recyclable polypropylene material.• The reel shall have the dimensions and capacities shown

in Table 8.4.

RecommendedMinimumCable Type/Fiber Count Grip Outer Diameter (in) Duct Size/Minimum Bend-Radius

12-36 Fibers 1.6 2.5-in with 18-in elbow48-144 Fibers 2.15 3.0-in with 18-in elbow

TABLE 8.3: Pulling Grip Specifications – MTP®-Terminated Trunks

Reel Capacities (ft)Trunk Fiber Count Reel A Reel B Reel C

12 5-99924 5-99936 5-99948 5-800 801-99972 5-450 451-99996 5-400 401-999144 5-300 301-800 801-999

Reel Dimensions (in)Reel A Reel B Reel C

Flange Diameter 23.5 23.5 23.5Drum Diameter 15.68 15.68 15.68TraverseWidth 5 12 18

TABLE 8.4: Reel Capacity and Dimensions


Harness Specifications and OptionsHarness Function and Construction• Harnesses shall be 12-fiber cable assemblies used as a

transition between MTP®-terminated trunk legs and endequipment ports or patch panels.

• Harnesses MTP Cable shall be plenum-rated.• The harness shall provide a means to transition from

MTP Connectors to LC duplex connectors. The break-out legs shall use a single two-fiber non-preferential bend2.0 mm cable terminated with duplex LC connectors andshare a single boot.

• The harness breakout point shall be a molded epoxy plug.• The harness epoxy plug shall include a feature that allows

mating two harnesses together in order to dress the fibersin an aesthetically pleasant manner. A hook-and-loopstrap shall be provided with every harness in order tosecure the harnesses together.

• Harness shall be color coded according to Table 8.5.• Specific breakout leg lengths and overall harness length

shall be tailored to meet the following equipment portlayout as indicated below.- Cisco 9513/9509/9506 LC stagger.- Brocade 48000/DCX-4S LC stagger- Cisco Nexus 7010/7018 LC stagger- Universal LC leg length of 6 in.

• The harness shall be available with 7 ft cable length(tail) for adjacent mounting of hardware and SANDirector and 10 ft cable length for flexible mountingoptions within the cabinet.

• Harness length shall be measured from the MTPConnector to the end of the furcation point.

Harness Fiber Types and Optical Specifications• Available fiber types and their optical performance

specifications shall be as indicated in Table 8.5.

Harness Connectivity• Harnesses shall be terminated with a pinned MTP

Connector and legs shall be terminated with duplexLC uniboot style connectors.

Jumper SpecificationJumper Function and Construction• The jumper shall be a 2-fiber cable assembly used as a

transition between the LC side of a harness or moduleand end equipment ports.

• Jumper shall be plenum-rated.• Jumper shall have LC duplex connectors and share a

single boot for both connectors.• The boot shall have an overall length from the connector

to the boot of 2.02 in.• The jumper shall be constructed with a single 2 mm

round cable with no preferential bend that allows easyrouting and reduces jumper congestion in the housingsand vertical managers.

Jumper Fiber Types and Optical Specifications.• Available fiber types and their optical performance

specifications shall be as indicated in Table 8.5.






MinimumOver Filled Launch(OFL) Bandwidth (MHz•km) 1500/500 3500/500 -/-


Jacket Color Aqua Aqua Yellow

Breakout Leg Colors Jacketed Same as Jacket

TABLE 8.5: Components Optical Specifications - Available Fiber Types, Colors

Notes:1) As predicted by RML BW, per TIA/EIA 455-204 and IEC 60793-1-41, for intermediate performance laser-based systems (up to 1 Gb/s).2) As predicted by minEMBc, per TIA/EIA 455-220 and IEC 60793-1-49 for high performance laser-based systems (up to 10 Gb/s).


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Adapter Panel SpecificationAdapter Panel Function and Construction• Panels shall meet the following dimensions

4.87 in x 3.53 in x 0.463 in (L x W x H).• Panels shall provide a means for joining MTP®-

terminated trunks entering the back of an MTP adapterpanel to a pinned MTP-terminated extender trunk orharness entering at the front of the panel.

• Panels shall be dimensionally compatible with CorningCable Systems LANscape® Pretium EDGE™ Solutionsrack-mountable connector housings.

• Panel design shall permit front and rear installation intothe Pretium EDGE Solutions housings.

Module Specifications and OptionsModule Function and Construction• Modules shall provide a means for joining MTP-

terminated trunks entering the back of an appropriatelydesigned connector housing to LC jumpers or cablesentering the front of the housing.

• Modules shall contain one 12-fiber cable assemblywithin a protective housing.

• Modules shall have shutter LC adapters at the front.• Modules shall be dimensionally compatible with Pretium

EDGE Solutions rack-mountable connector housings.• The small form module shall meet the following

dimensions 4.87 in x 3.53 in x 0.463 in (L x W x H).It shall provide a high-density solution when loaded intothe 01U and 04U Pretium EDGE Solutions housings.

• Modules shall permit front and rear installation intothe Pretium EDGE Solutions housings.

• When uninstalling a module from the back, a rearaccessible retention trigger and finger handle must bepresent in order to facilitate this operation. An I.D.and warranty seal label shall be affixed to every module.

• When mounted in a connector housing, the adaptersleeves shall be accessible from the front, thus providinga cross-connection point with other modules.

• Modules shall contain discrete fiber and portidentification. This fiber and port identification shallbe pad printed on top and bottom of the modules.

Insertion Loss, max (dB)*


Priority Pretium 300Ultra-BendableOptimized 50µm (850/1300nm)



MTPMated Pair Loss 0.35 0.35 0.75

LC Mated Pair Loss 0.15 0.15 0.5

Module Loss 0.5 0.5 1.3

TABLE 8.7: Components Optical Specifications - Available Fiber Types

*Insertion loss specifications when mated to other system components of a like performance specification.






MinimumOver Filled Launch(OFL) Bandwidth (MHz•km) 1500/500 3500/500 -/-


Adapter ColorLCMTP

AquaAqua

AquaAqua

BlueBlack

TABLE 8.6: Modules - Available Fiber Types, Optical Specifications, Adapter Colors

Notes:1) As predicted by RML BW, per TIA/EIA 455-204 and IEC 60793-1-41, for intermediate performance laser-based systems (up to 1 Gb/s).2) As predicted by minEMBc, per TIA/EIA 455-220 and IEC 60793-1-49 for high-performance laser-based systems (up to 10 Gb/s).


Module Connectivity• Cable assemblies within modules shall be terminated with

MTP® Pinned Connector at the back and LC connectorat the front.

• Each module shall contain 12 fiber terminations.• All connectors shall be inside the module but shall be

accessible for mating through adapter sleeves mountedthrough the wall of the module.

• Module shall have a self-retracting shutter adaptermechanism that allows a single hand operation.

• The shutter adapter shall eliminate the need to removeand re-install dust caps. The shutter adapter shall be VFLcompatible. The adapter sleeves shall be color coded asindicated in Table 8.6.

Module Fiber Types and Optical Specifications• Available fiber types and their optical performance

specifications shall be as indicated in Table 8.6.• Module insertion loss performance shall be as indicated

in Table 8.7.

Module packaging• The modules shall be packaged in blister packs.

The blister pack’s overall dimensions shall be4 3/4 in x 3/4 in x 7 3/16 in.

• The blister packs shall have the ability to be storedin a box or be hung when using hook merchandisingstorage device.

• The plastic reel shall be constructed with 100 percentrecyclable polyethylene terephthalate (PET) material.

Components Insertion Loss SpecificationsAll components shall meet the maximum insertion lossvalues indicated in Table 8.7.

Universal Polarity Management SystemTrunks, modules and harnesses shall follow the fiberrouting schematic of Figure 8.2 and Figure 8.3.• Standard ribbon position is defined as having the end

face of the blue fiber on the left of the MTP Connectoras the MTP end face is viewed in the key-up position.

• Reverse ribbon position is defined as having the end faceof the blue fiber on the right of the MTP Connector asthe MTP end face is viewed in the key-up position.

• Keys schematically represented in the down position aredrawn with a dashed line.

• All MTP Connectors shall mate key-up to key-down.• Primary trunks shall have MTP Connectors on one end

oriented in the standard ribbon position and MTPConnectors on the other end oriented in the reverseribbon position.

• Extender trunks shall have both MTP Connectorsinstalled in the Standard Ribbon Position.

• Modules and harnesses shall contain MTP Connectorsin the Standard Ribbon Position.

• Modules shall have polarity-managed fiber routing asshown in Figure 8.2.

• Harnesses shall have polarity-managed fiber routingwithin a furcation plug as shown in Figure 8.3.

IdenticalUniversalModules

GuidePin

Simplex/Duplex Fiber

MTP Key-upwith Reverse

RibbonPositioning

MTPKey-down

Z

Figure 8.2Universal Wiring Scheme – Modules on Both Ends| Drawing ZA-3591

Simplex/Duplex Fiber Terminations

MTP Key-upwith Reverse

RibbonPositioning

MTPKey-down

GuidePin

UniversalModule

UniversalHarness

FurcationPlug

Figure 8.3Universal Wiring Scheme – Harness on One End| Drawing ZA-3592

Notes:1) All MTP Connectors shall be installed in standard ribbon positionexcept as noted.2) Extender trunk shown in upper right corner is optional.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Generic Specification for 1U Pretium EDGE™

Solutions HousingRack-Mountable Connector HousingsRack-mountable connector housings shall be available forcross-connecting or interconnecting purposes.

Standards• Housings shall be mountable in an EIA-310 compatible

465 mm (18.3 in) rack. One EIA rack space or panelheight (denoted as 1U) is defined as being 44.45 mm(1.75 in) in height.

1U Housing• Housings shall be available in a 1U size.• The housing shall be modular, allowing the installation

of 12-fiber Pretium EDGE™ modules in order to providescalability in increments of 12 fibers. The maximumhousing density shall be 96 fibers when it is fully loadedwith modules.

• The unit shall be mounted with a 5.33 infrontal projection.

• The unit shall not exceed a depth requirementof 16.3 in.

• The 1U Pretium EDGE Solutions housing shall havetwo sliding trays contained in a single drawer and shallallow the installation of four modules per tray. Thedrawer shall slide out and tilt 25 degrees for easymodule installation.

• The unit shall meet the design requirements ofANSI/TIA/EIA-568 and the plastics flammabilityrequirements of UL 94 V-0.

• Housings shall be manufactured using 0.050 in aluminumor equivalent for structural integrity. The housing’sdrawer and mounting brackets shall be manufacturedwith 18-gauge cold rolled steel. The housing shall befinished with a reflective silver coat for durability.Installation fasteners shall be included and shall beblack in color.

Tray• The 1U Pretium EDGE Solutions housing shall have

two sliding trays and each having four modules capacity.• Each tray shall provide connectivity through 48 LC

connectors when fully loaded.• Each individual tray shall have patch cord routing guides

that allow a transition and jumper management point.The jumpers shall be able to exit through the right andleft sides of the housing. This jumper managementscheme shall provide access to individual trays to easeadministration in high-density applications.

• The trays shall be manufactured using 18-gauge coldrolled steel or equivalent for structural integrity andshall be finished with reflective silver powder coat fordurability.

• The trays shall slide 3.6 in to the front in order toprovide appropriate finger access to the connectors andmodules. The tray shall have a closed and open positionwith their respective mechanical stops.

• The Pretium EDGE Solutions hardware shall provideaccess to each adapter port with no interference ofadjacent ports. In addition the accessibility to theconnectors shall be tool-less.

• The trays shall have a cut-out in front of each modulelocation in order to provide accessibility from the topand bottom of adapters and modules.

• The trays shall incorporate rails to facilitate front andrear module installation while providing a lockingmechanism that secures the module in place.

• The trays shall have protruding finger tabs on the sidesto allow easy access to modules and connectors. Thetabs shall have silk screened numbers for tray identification.

• The rails shall incorporate a release button, engravedwith the word “push”, which allows removal of modulesfrom the front.

• The trays shall provide visible module identificationwith the letters A, B, C and D.

The unit shall have eight trunk strain-relief locationswhich allow fully loading the housing to its maximumcapacity using 12-fiber trunks.

The housing shall contain a front door. This door shall behinged with a pivot point at the bottom of the housing.The door shall utilize a sliding latch mechanism to provideeasy access when opening and closing.

The housings shall have a removable cover at the backof the housing and shall provide protection to trunk legs.The connector housings shall have a labeling scheme thatcomplies with ANSI/TIA/EIA-606.


Generic Specification for 4U Pretium EDGE™

Solutions HousingRack-Mountable Connector HousingsRack-mountable connector housings shall be available forcross-connecting or interconnecting purposes.

Standards• Housings shall be mountable in an EIA-310 compatible

465 mm (18.3 in) rack. One EIA rack space or panelheight (denoted as 1U) is defined as being 44.45 mm(1.75 in) in height.

4U Housing• Housings shall be available in a 4U size.• The housing shall be modular, allowing the installation

of 12 fiber PretiumEDGE™ Solutions modules inorder to provide scalability in increments of 12 fibers.

• The maximum housing density shall be 576 fibers whenfully loaded with modules.

• The unit shall be mounted with a 5.33-in frontalprojection.

• The unit shall not exceed a depth requirement of 18.35-in.• The unit shall meet the design requirements of

ANSI/TIA/EIA-568 and the plastics flammabilityrequirements of UL 94 V-0.

• Housings shall be manufactured using 0.063-inaluminum or equivalent for structural integrityand shall be finished with a reflective silver powdercoat for durability. Installation fasteners shall beincluded and shall be black in color.

• The housing shall include two field-installable slackmanagement brackets at the front of the housing.The brackets shall provide jumper slack managementat the front of the housing and shall allow easy traydeployment when the tray is fully loaded with PretiumEDGE Solutions jumpers.

Tray• The 4U Pretium EDGE Solutions housing shall have

12 sliding trays with each having a four modules capacity.• Each tray shall provide connectivity through 48 LC

connectors when fully loaded.• Each individual tray shall have patch cord routing guides

that allow a transition and jumper management point.The jumpers shall be able to exit through the right andleft sides of the housing. This jumper managementscheme shall provide access to individual trays to easeadministration in high-density applications.

• The trays shall be manufactured using 18-gauge coldrolled steel or equivalent for structural integrity andshall be finished with reflective silver powder coat fordurability.

• The trays shall slide 3.6-in to the front in order toprovide appropriate finger access to the connectors andmodules. The tray shall have a closed and open positionwith their respective mechanical stops.

• The Pretium EDGE Solutions hardware shall provideaccess to each adapter port with no interference ofadjacent ports. In addition, the accessibility to theconnectors shall be tool-less.

• The trays shall have a cutout in front of each modulelocation in order to provide accessibility from top andbottom of adapters and modules.

• The trays shall incorporate rails to facilitate front andrear module installation while providing a lockingmechanism that secures the module in place.

• The trays shall have protruding finger tabs on the sidesto allow easy access to modules and connectors. Thetabs shall have silk screened numbers for tray identification.

• The rails shall incorporate a release button, engravedwith the word “push”, which allows removal of modulesfrom the front.

• The trays shall provide visible module identificationwith the letters A, B, C and D.

The unit shall have 24 trunk strain-relief locationsallowing fully loading the housing to its maximum capacityusing any trunk fiber count. When deploying 12-fibertrunks, a double stack strain-relief method shall be used.

The housing shall incorporate three strap points in orderto secure the trunk legs with hook-loop straps preventingthese from exiting the housing perimeter.

The housing shall contain front and rear doors. Thesedoors shall be hinged with a pivot point at the bottom of thehousing. The doors shall utilize a sliding latch mechanismto provide easy access when opening and closing.

The housings shall have two open-ended slots, one oneach side for quick and easy trunk installation. These slotsshall be covered by cable entry brushes.

The connector housings shall have a labeling scheme thatcomplies with ANSI/TIA/EIA-606.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

An important step in data center design is procuring theproduct. Shipping times will affect the overall scope ofthe project. Planning the logistics in a data center buildwill depend on product procurement. Corning CableSystems products are widely distributed and can bepurchased from the following distribution channels:

When you demand only the best from your network,you can depend on Accu-Tech. Accu-Tech is stronglycommitted to providing superior products and support todeliver unparalleled business experience to their customers.

www.accu-tech.com

Anixter is a leading global supplier of communications andsecurity products, electrical and electronic wire and cable,fasteners and other small components. Anixter helps theircustomers specify solutions and make informed purchasingdecisions around technology, applications and relevantstandards. Throughout the world, Anixter provides inno-vative supply chain management services to reduce theircustomers’ total cost of production and implementation.

www.anixter.com

Founded in 1972, Communications Supply Corporation isa leading distributor of low-voltage network infrastructureand industrial wire and cable products.

Through a network of 33 branch offices, CSC distributesa full range of products to support advanced connectivityfor voice and data communications, access control, securitysurveillance, building automation, video distribution, lifesafety broadcast systems and electrical construction andmanufacturing for commercial, residential and governmentcustomers.

CSC is recognized for delivering measurable value andoutstanding support to its customers and suppliers alike.Vast application expertise makes CSC an unbiased knowl-edge resource for product information, documentationand training.

www.gocsc.com

Graybar has specialized in supply chain managementservices and distribution of high-quality components,equipment and materials for the electrical andtelecommunications industries for over 80 years.

Incorporated in 1925, Graybar procures, warehouses, anddelivers just about any kind of electrical or communica-tions and data product, component, or related service toits customers. It stocks and sells hundreds of thousands ofitems from thousands of manufacturers.

www.graybar.com

Chapter Nine:Procuring the Data Center Products

Chapter Nine: Procuring the Data Center Products | LAN-1160-EN | Page 47

Choosing the ContractorA key decision in the data center network is choosing acontractor for the installation. Some key questions thatneed to be addressed before selecting a contractor are:1. How long has the contractor been in business?2. Is structured cabling the contractor’s core business?3. What percentage of the contractor’s business is from

structured cabling?4. Does the contractor install fiber optic cable

and hardware?5. Does the contractor have a market expertise in

data centers or server farms?6. Does the contractor have adequate insurance

and bonding?7. Does the contractor belong to professional

organizations, such as BICSI?8. Is the contractor certified by the manufacturers?

Is that certification current?9. Does the contractor have approvals and licenses from

unions as well as safety and construction boards?10.Which manufacturers does the contractor represent?

The fiber optic contractor should be able to work withthe customer in each installation project through thesethree key areas:1. Network design2. Installation3. Post installation

• Testing, troubleshooting, documentation,restoration

The contractor should be experienced in fiber opticinstallations and should provide references.

Network DesignA good contractor should be able to assist with thedesign process. The contractor should be able tohelp the customer:1. Choose the correct optical fibers

- OM2, OM3, OM4, OS22. Choose the correct optical cables

- Outside plant, indoor, riser, plenum3. Choose the correct hardware

- High-density, connector type4. Choose the correct vendors5. Understand standards

- TIA-942

InstallationA good contractor should also be able to assist with theinstallation process. The contractor should be able tohelp the customer:1. Purchase, receive, inspect and bring components

to the work site2. Choose components that they have been trained

to install

The technicians actually doing the installation should betrained and certified by manufacturers of the productsbeing installed or by different organizations such as:1. BICSI2. FOA (Fiber Optic Association)

Certification is important and protects the end user.

Post Installation: Testing, Troubleshooting,Documentation and RestorationAll of these items need to be discussed with the contractorbefore work begins. The contractor and customer needto agree on what is covered in the scope of work. Forexample, testing may be included, but troubleshootingand restoration may not be included. It is good to havea clear understanding with the contractor on what itemswill be covered.

Chapter Ten:Installation

Chapter Ten: Installation | LAN-1160-EN | Page 48

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

What is Polarity andWhy Does it Matter?Polarity is the process of ensuring that the information sentby a transceiver on the transmitter (Tx) port is received bythe end equipment on the receiver (Rx) port. The primarygoal of polarity in the backbone infrastructure is to ensurethat in every panel, the Tx port ends up at an Rx port.This ensures that when craft connects your end equipment,they will simply connect Tx on the panel to Rx on theequipment, and Rx on the panel to Tx on the equipment.If polarity is not actively managed in the backbone, theonly way to install the patch cabling would be through trialand error. If there are multiple patching areas in a system,this becomes a very difficult task.

To ensure that polarity is maintained in a structuredcabling environment, TIA/EIA 568-C.0 describesthree sample methods that may be used to manage thischallenge. The standard is very specific in distinguishingbetween mandatory items and suggested items. Polarityis an area where the standard recognizes that a varietyof methods may be employed and that listing them allwould not be possible. As such, they listed three samplemethods that we will discuss, and they left open theoption to improve upon those methods while remainingstandards-compliant.

Simplex and Duplex Connector PolaritySimplex and duplex connectors and adapters are all keyedto ensure the same orientation of the connector upon mat-ing. This keying establishes the orientation of one fiber tothe other (polarity) because the simplex/duplex connectorscan only insert into the adapter in one direction. Polarity ismanaged at the patch panel or outlet by using consecutivefiber numbering and rotating the adapter on one end ofthe link or by installing backplane fibers using reverse-pairpositioning. Polarity is important so “transmit” is notaccidentally plugged into “transmit.” Both polarity methodsare accepted by TIA/EIA-568-C.0 and C.3.

Figure 11.1 shows the difference in fiber configurationsfor single-fiber and single-ferrule duplex connectors.Optical fiber cabling should be installed so that theodd-numbered fiber within the cable is paired with thenext consecutive even-numbered fiber (e.g., fiber 1-bluepaired with fiber 2-orange, fiber 3-green paired withfiber 4-brown). This installation of paired fibers formsthe Tx and the Rx transmission paths used in providinga telecommunications circuit.

To achieve these transmission paths while retaining cableintegrity, reverse-pair positioning may be used. Reverse-pair positioning is achieved by installing fibers in consecu-tive numbering sequence (i.e., 1, 2, 3, 4 …) on one end of

KeySC Connectors

Duplex Clip

Blue Fiber (A)

Single-Fiber Cables

Orange Fiber (B)

Orange Fiber (B) Blue Fiber (A)

Key(top of the ferrule)

MT-RJ Ferrule

Fiber

Cable Jacket

Figure 11.1Simplex and Duplex Connector Polarity | Drawing ZA-2418

Chapter Eleven:Polarity

Chapter Eleven: Polarity | LAN-1160-EN | Page 49

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Performance Metrics and Administration 4SECTION

an optical fiber link and installing fibers using reverse-pair numbering (i.e., 2, 1, 4, 3 …) on the other end of theoptical fiber link.

Optical fiber patch cords, as specified in TIA/EIA-568-C.3,must be used when completing circuits using reverse-pairpositioning.

Implementing Reverse-Pair PositioningTo implement reverse-pair positioning in the cablingsystem, the following steps should be taken.

1. Assign each fiber in a given cable a sequential numberfollowing the same order as described in TIA/EIA-568-C.3 (see Table 11.1).

2. Install connectors on both ends of the cable as follows(see Figure 11.2):

a) On one end of each cable, install the fibers inconsecutive order (i.e., 1, 2, 3, 4 …).

b) On the other end of each cable, install the fibers inreverse-pair numbering (i.e., 2, 1, 4, 3 …).

Notes:1. From the installer’s point of view, fiber 1-blue will appear on the left on one

end and on the right on the other end of every link. Fiber 2-orange willappear in the opposite manner, right on one end of the link and left on theother end of the link.

2. Reverse-pair positioning may be obtained by installing the fibers on theconnectors in this manner or by installing connectors into the adapters inthis orientation.

3. Successive cables placed in the channel (e.g., MC to IC, IC to HC) shouldbe installed as described above.

Array Connector PolarityDense data center wiring requirements dictate the use ofarray-style connectors like the MTP® Connector. Thesescenarios often utilize factory-terminated MTP to MTPconnectorized cables or trunks. Since there are arrayconnectors on both ends of these trunks, and the end

equipment typically has standard duplex transceiver ports,the trunks are plugged into a factory-made furcation ormodule that transitions from the MTP Connector to aduplex connector/adapter style. Like simplex and duplexconnectors and adapters, the MTP Connectors andadapters are also keyed to ensure the proper orientationis maintained when connectors are mated. With MTPConnectors, this keying establishes the orientation of onefiber array in one connector relative to the array in themating connector, but does not ensure that fiber-pairpolarity is maintained. This is accomplished in one ofseveral different methods. These methods are examinedin the following diagrams.

ICMC HC WA

Rx Tx Rx Tx

Front

Back

Blue Backbone FiberOrange Backbone Fiber

Blue Patch Cord FiberOrange Patch Cord Fiber

Legend:

Figure 11.2Reverse-Pair Positioning | Drawing ZA-2419

TABLE 11.1: Polarity

Fiber Number Color Fiber Number Color

1 Blue 7 Red2 Orange 8 Black3 Green 9 Yellow4 Brown 10 Violet5 Slate 11 Rose6 White 12 Aqua


Method AMethod A (Figure 11.3) uses a single module type wiredin a straight-through configuration and two different patchcords in an optical circuit. One patch cord is straight wiredand the other has a pair-wise flip. All components in thechannel are mated key-up to key-down. No guidance isincluded in the standard to differentiate where the patchcord with pair-wise flips should be used, and how to makeit easily recognizable from the regular straight-wiredduplex patch cord. Because polarity is addressed in thepatch cords, the end user is ultimately responsible formanaging it.

Method BMethod B (Figure 11.4) uses a single module type wired ina straight-through configuration and standard patch cordson both ends. The differences are that all components inthe system are mated key-up to key-up. When the link isconfigured in this fashion, physical position #1 goes tophysical position #12 on the other end. A module on oneend is inverted so that logically (labelwise), position #1goes to position #1. This method requires advanced plan-ning for module locations in order to identify the moduletypes and location of the inverted module in the opticallink. This adds complexity to the polarity management.Using an MTP® Connector key-up to key-up configurationdoes not easily accommodate angled polished (APC)single-mode connectors.

Method CMethod C (Figure 11.5) uses a pair-wise flip in thetrunk cable to correct for polarity. This enables the useof the same module type on both ends of the channel andstandard patch cords. Because polarity is managed in thetrunk, extending the links requires planning of the numberof trunks in order to maintain polarity. The TIA standarddoes not include text regarding the ability to migrate toparallel optics for Method C, but parallel optic capabilitycan easily be achieved with a special patch cord to reversethe pair-wise flips in the trunk.

Universal Polarity Management MethodThe Universal Polarity Management Method (Figure 11.6)is an enhanced polarity management method that improvesupon the sample methods listed in TIA/EIA 568-C.0 aswas the intent of the standard. The method uses the samemodule and patch cord type at both ends with no inversionor reconfiguration needed to maintain polarity. Polarityis easily accomplished and managed with the modules’


Figure 11.5Method C | Drawing ZA-3028

Figure 11.3Method A | Drawing ZA-3026

Figure 11.4Method B | Drawing ZA-3027

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

internal fiber wiring scheme. The system is mated key-upto key-down. The method supports simple concatenationof multiple trunks without affecting polarity. The methodeasily accommodates all simplex/duplex connector types aswell as single-mode fiber APC MTP® Connectors. Similarto Methods A, B and C, the universal polarity managementmethod easily facilitates migration to parallel optics. Thewired modular system components enable fast and simplenetworking moves, adds and changes without polarityconcerns associated with special polarity-compensatingcomponents used in Methods A, B and C.

Parallel Optics for 40G and 100G EthernetThe IEEE standard for 40G and 100G Ethernetemploys a parallel optics scheme for multimode fiber.The IEEE 802.3ba task force has specified paralleltransmission of 40G and 100G Ethernet to 100 m withOM3 fiber and 125 m with OM4 fiber. 40G Ethernetparallel optics transmission will utilize the current 12-fiberMTP Connector while 100G Ethernet will utilize a 24-fiberMTP Connector to transmit data on multiple fibers.

For example, for 40G Ethernet, eight fibers from a 12-fiberMTP connector would be used. Four fibers would be usedfor Tx at 10G, and four others would be used for Rx at10G for an aggregate signal of 40G. To transmit 100GEthernet, one would use 20 fibers on a single 24-fiberMTP Connector (Figures 11.7 and 11.8). Ten wouldtransmit 10G each or 100G in aggregate. The other10 would then receive 10G each or 100G in aggregate.The end electronics would then multiplex the data.Corning Cable Systems’ Universal Polarity ManagementMethod is fully compatible with the final polarity schemedeveloped by this task force.

Trunk with Standard MTP andRibbon Twist MTP Connectors

Module withMTP Connector

Module withMTP Connector

R L R L R L R L R L R L R L R L R L R L R L R L✦✦

Key-Up to Key-Down Key-Up to Key-Down

Figure 11.6Universal Polarity Management Method | Drawing ZA-3486


RxRxRxRx

TxTxTxTx

TxTxTxTx

RxRxRxRx

Fiber Position12 1

Fiber Position 121



Optical ReceiverMTP Connector


Z



Testing of any installed cabling system in the data centeris crucial to ensuring the overall integrity and long-termperformance of the network. Documenting test resultsquantifies system quality, identifies system faults and estab-lishes accountability when multiple vendors are involved.Simple, reliable and field-proven test procedures are alreadyestablished for certifying that an optical fiber cabling systemis properly installed. Proper testing also maximizes thesystem’s longevity, minimizes downtime and maintenance,and facilitates system upgrades or reconfigurations.

This chapter addresses testing, documentation and mainte-nance of optical fiber cabling systems for new installations,system upgrades and individual components in the data cen-ter. With more than 25 years of field experience, CorningCable Systems offers straightforward test procedures andpractical guidelines for system testing. This informationis in accordance with TIA/EIA-568-C. Since the standardaddresses only the end-to-end attenuation test, we havecombined this with additional information on other testmethods and common field test practice applicable in thedata center.

Attenuation, defined as optical power loss measured indecibels (dB), is the primary test parameter in optical fibersystems. Cables, connectors, splices and patch cords allcontribute to the system’s overall attenuation. Additionalloss may also be induced by tight bends or excessive forcesplaced on the cable during transport and installation.Testing must be done after installation to ensure that thecable system meets the attenuation specifications set forthby the end user. Implementing the recommendations of thischapter provides solid proof of system integrity and ensuresreliable system operation.

Cable System TestingEnd-to-end attenuation and OTDR tests provide quantita-tive measures of the installed performance of the cablesystem and its components. This section outlines the basicconcepts, test methods, test equipment and specific appli-cation guidelines for each type of testing. A summary ofrecommendations for cable system testing by segmentconcludes the section.

In preparing for data center tests, the following guidelinesare important for efficient and accurate test results:

• Ensure that the test jumpers (end-to-end attenuation)or test fiber box (OTDR) are of the same fiber core sizeand connector type as the cable system, e.g., 50/125 µmcore test jumpers should be used for testing a 50/125 µmmultimode cable.

• Ensure that optical sources are stabilized and havecenter wavelengths within ± 20 nm of the 850/1300 nmmultimode and 1310/1550 nm single-mode nominalwavelengths. In accordance with TIA/EIA-526-14-A,multimode LED sources should have spectral widthsfrom 30-60 nm at 850 nm and 100-140 nm at 1300 nm.

• Ensure that the power meter and the light source are setto the same wavelength.

• Ensure that all system connectors, adapters and jumpersare properly cleaned prior to and during measurement.

End-to-End Attenuation TestingThe single most important test of an installed link is end-to-end attenuation. This is a measure of the optical powerloss between cable termination points. Acceptable lossvalues are dependent upon the system length, wavelengthand number and type of connectors and splices. The end-to-end loss should always be less than the link-loss budgetcalculated in the system design. The best way to verify thatthe cable meets the loss limit is to measure each segmentfrom patch panel to patch panel. Because of the stress andbending that cables can be subjected to during installation,Corning Cable Systems recommends measuring theattenuation of each connectorized link after installation.

Chapter Twelve:Testing and Documentation

Chapter Twelve: Testing and Documentation | LAN-1160-EN | Page 53

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

BackgroundThe attenuation of installed cable systems is measuredby the insertion loss method. This method uses an opticalsource and optical power meter to compare the differencebetween two optical power levels – first measuring howmuch light is put into the cable at the near end, and thenmeasuring how much light exits the far end after the cablesystem is inserted in between.

These absolute optical power levels are measured in dBm.By definition, dBm = 10 log (Pout)/1 mW of power, hencethe “m” in dBm.

Loss (dB) = P2 (dBm) - P1 (dBm)Where: P2 = Output Power (dBm)P1 = Input Power (dBm)

ProcedureEnd-to-end attenuation testing is performed by a simplethree-step procedure in accordance with TIA/EIAspecifications:• Multimode fiber: OFSTP-14A

• Single-mode fiber: OFSTP-7A

The procedure described here is for patch panel to patchpanel applications only.

A stabilized light source and optical power meter are usedto quickly and accurately measure the attenuation of eachterminated fiber as shown in Figure 12.1. Best results areachieved with factory-terminated test jumpers.

End-to-End Attenuation Test forSingle-Fiber ConnectorsStep 1: ReferenceFor the TIA/EIA-568-C compliant networks, Corning CableSystems recommends the use of a 1-jumper reference asdescribed in the procedures below. Performing the 1-jumperreference provides the most accurate and appropriate testfor your system. Additional jumper referencing will falselyimprove results by eliminating potential loss events.Note: A 2-jumper reference should only be used when your system begins ata patch panel and ends directly in end equipment. Additionally, a 3-jumperreference should only be used when your system begins and ends directly in theend equipment.

Connect a short test jumper between the optical sourceand the optical meter. Ensure that the reference power indBm is within acceptable range per unit specification. Thispower level is simply the output power of the light sourcecoupled into the jumper to the meter (See Figure 12.1).Press the reference button on the meter, and the meterreading should then read 0dB. Note: Never disconnect or adjustthe jumper connection at the optical source after recording the reference value.

This can change the value.

Mandrel WrappingIn accordance with TIA/EIA-568-C.1, mandrel wrappingshould be used when performing power through testingon multimode fiber. Optical fibers are designed to attenu-ate the cladding modes almost immediately. Along withthe light in the core, there may be some high-order modesin the cladding due to the fully flooded launch condition.These high-order modes normally have a much higherattenuation than lower-order modes, and often will notappear at the far end of a fiber link of sufficient length.Due to these high-order modes, issues arise during thereferencing step of a typical attenuation test.

Referencing occurs with test jumpers that are only a fewmeters in length. Over a short distance, the high-ordermodes do not completely dissipate before reaching thetest meter. This extra optical power is calculated into thereference. When actual system testing occurs, however,the higher-order modes completely dissipate over thelength of the system and do not reach the meter. Thisdifference in power gives the appearance of a higher-losssystem. To prevent the high-order modes from invalidatingthe test results, they need to be attenuated during thereferencing step to obtain a valid measure of the opticalpower that will actually travel along the fiber core. This isoften done by wrapping a length of fiber around a smooth,round mandrel (rod) during the testing process. The fiberused should be long enough to allow for five wraps around

OTS-600

1

4

8

0

7 9

5 6

2 3

Test Jumper #1

Power Received = Preference (dBm)= -20.0 dBm

-20.0 dBm

Five Turns Around a Mandrel(Multimode Fiber Only)

F

OTS-600

1

4

8

0

7 9

5 6

2 3

Figure 12.1End-to-End Attenuation Test | Drawing ZA-3593


Figure 12.2OTS-600 Series Optical Source and Meter | Photo LAN1199

OTS-600

1

4

8

0

7 9

5 6

2 3

0.4 dBPower Received = P = 0.4 dBcheck

Test Jumper #2

Adapter(System Connector Type)

Test Jumper #1



O

OTS-600

1

4

8

0

7 9

5 6

2 3



INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

the mandrel. The bending caused by wrapping the fiberaround the mandrel will strip out (attenuate) the high-order modes in the cladding. User Tip: Although the output

connector of the optical source can be different from the system connector type, the

optical meter’s connector input must match the system. It is important, therefore,

that the optical meter have interchangeable connector adapters. Additionally,

pure optical test equipment rather than copper test equipment with optical

capabilities is recommended for the most reliable and accurate results.

Figure 12.1 is an example of a test setup that incorporatesa mandrel wrap. Figure 12.2 shows the step of referencingthe optical power output of the test source and then thetesting of a system. Note that at the transmitter, themandrel and wrapped jumper are used both during thereference step and during the system testing.

Refer to TIA/EIA-568-C.1 or Corning Cable SystemsApplication Engineering Note 68 on mandrel wrappingduring multimode testing.

Step 2: CheckDisconnect test jumper no. 1 from the power meter andinsert a second test jumper (test jumper no. 2), using anadapter, between the jumper used in Step 1 and the opticalpower meter. Verify that the two test jumpers are good byensuring that the power is within the appropriate connec-tor loss, typically < 0.5 dB. If this criterion is met, contin-ue to Step 3. Note: Do not reference at this point.

If the criterion is not met, clean all connectors except thesource connection point and repeat Step 2. If the loss isstill greater than 0.5 dB, replace test jumper no. 2 andrepeat Step 2. If the loss is still greater than 0.5 dB, tryreplacing the adapter and repeat Step 2 (see Figure 12.3).

Mandrel Diameter Mandrel DiameterFiber Core Size For 3mm (0.12 in) cable For 2mm (0.08 in) cable

50 µm (Standard OM2/OM3) 22 mm (0.87 in) N/A50 µm (Corning® Ultra-Bend) @ 850 nm N/A 4 mm (0.16 in)50 µm (Corning Ultra-Bend) @ 1300 nm N/A 22 mm (0.87 in)62.5 µm 17 mm (0.67 in) N/A

TABLE 12.1: MandrelWrapping with Multimode Fiber

Note: Mandrel part numbers currently available in NAFTA and ready for shipment.OTS-MANDREL-50OTS-MANDREL-62OTS-MANDREL-4OTS-COMBOMAN (This part number comes with the Standard 50 µm and 62.5 µm.

Step 3: TestLeave the two test jumpers attached to the optical sourceand optical meter. Disconnect the two jumpers at theadapter. Attach the optical source/test jumper no. 1 toone end of the system fiber to be tested and the powermeter/test jumper no. 2 to the other end of the samefiber. Record the losses for each fiber to be tested(see Figure 12.4 and 12.5).

End-to-End Attenuation Test forMTP® Pinless Connector LinksEquipment required for this test:• Optical source with SC optical port

• Optical meter with SC optical port

• 12-fiber SC to MTP pinned Connector hybridjumpers – two

• 12-fiber MTP Connector to MTP Pinless Connectorjumper – one

• SC-SC jumper – three

• MTP Connector adapters – two

• SC adapters – two

Note: The example herein utilizes a light source and power meter that eachhave an SC connector interface; other single-fiber interface types work in asimilar fashion.

OTS-600

1

4 5 6

2 3

-7.7 dB850nm

Figure 12.5OTS Display | Drawing ZA-3594

System

Power Received = P = 7.7 dBcheck

T

Power Received = P test

PatchPanel

PatchPanel

TestJumper

#2

TestJumper

#1

ZA-3593

F


OTS-600

1

4

8

0

7 9

5 6

2 3

OTS-600

1

4

8

0

7 9

5 6

2 3

7.7 dB



SC ConnecMTP Pinned Connector

M

Blue Leg Blue Legectors

M

Connector Pair

Connector Pair

SC Jumper No. 1

SC Jumper No. 3

SC Jumper No. 2Source

Meter-19.0 dBm

O

Figure 12.8Setting up the Reference Step with Three Jumpers| Drawing ZA-3596

Source

SC ConnectorsMTP Pinned Connector

MTP Pinless Connector

Blue Leg Blue Leg

Aqua LegAqua Leg

SC Connectors

Meter1.5 dBm

SC Connectors

MTP Pinned Connector


L

Bl L Bl LSC Connectors

OTS-600

1

4

8

0

7 9

5 6

2 3

OTS-600

1

4

8

0

7 9

5 6

2 3

Figure 12.9Setup and Verification of Test Jumpers | Drawing ZA-3596

SC Jumper No. 1

Source Meter-18.0 dBm

SC Jumper No. 1 SC Jumper No. 2

Connector pair

S

Figure 12.6Determining the Output Power of the Source Using OneJumper | Drawing ZA-3596

SC Jumper No. 1 SC Jumper No. 2

Connector pair

Source Meter-18.5 dBm

Connector Pair

C

SC Jumper No. 1

SC Jumper No 3

Figure 12.7Checking the Test Connectors | Drawing ZA-3596

Step 1: Setting the Reference(using the 3-Jumper Reference Method)Connect the ends of the SC jumper to the optical sourceand meter as shown in Figure 12.6. Ensure that the powerreading in dBm is within the specified range of the opticalsource for the fiber type under test. Press the referencebutton on the meter.

Insert a second SC jumper into the setup as shown inFigure 12.7, connecting to test jumper no. 1 on one endand to the meter on the other. The loss reading shouldnot be higher than the value specified for the test jumperconnectors, typically 0.5 dB or less for factory-terminatedsingle-fiber connectors. Press the reference button onthe meter.

Decouple the connector pair made in the previous step.Insert test jumper no. 3 between jumper no. 1 and no. 2as shown in Figure 12.8. The loss reading should notbe higher than 0.5 dB. If a higher than expected loss ismeasured, clean the connectors and retest. If the jumperscontinue to test high, replace each jumper with a new oneuntil the measurement reading is in the appropriate range.Press the reference button on the meter. The meter shoulddisplay 0.0 dB.

Note: Contrary to earlier guidance, a 3-jumper reference should be usedfor an MTP®-MTP Connector link due to the cable/system configuration.MTP Connector links are typically terminated in multi-fiber connector modulesor directly into end-equipment requiring the use of a 3-jumper reference.

Step 2: Checking MTP Connector Test JumpersRemove jumper number 3 from the test setup. Connectthe blue leg of a 12-fiber SC to MTP Pinned Connectorhybrid jumper to the SC jumper at the source and theblue leg of a second 12-fiber SC to MTP Pinned HybridConnector jumper to the SC jumper at the meter, asshown in Figure 12.9. Connect the test sets and testjumpers together with an MTP Connector to MTPConnector (both WITHOUT pins) jumper. For properpolarity testing with standard jumpers, the same jumperleg (same number or color) must be connected to thesetup for each measurement.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

The meter should now display a negative value of (≤ 1.0 dB).Do Not Reference Here (do not press the reference buttonon the meter). These values are obtained by using the maxi-mum loss of 0.5 dB for a single-mated MTP® Connectorpair. This value can be taken from the manufacturer’sspecification for maximum connector pair loss. The maxi-mum resultant sum of two mated pairs would be a 1.0 dBloss. Disconnect the blue leg of each SC to MTP PinnedHybrid Connector jumper and connect with the orange legs.Test through all 12 SC connectors in sequence, ensuring thatall connectors involved in the testing process are sound; eachreading should be below the acceptable level. After verifyingall 12 SC legs, remove the pinless MTP Connector toMTP Connector jumper from the setup.

The system is now ready to test.

Step 3: TestWithout disconnecting from the units, take source and meterto the distant ends of the system (Figure 12.10). Each testvalue represents the system loss along one run of fiber.

Reconnect the first SC connector of each MTP Connectorto SC cable assembly to the source and meter SC jumpers.Connect the MTP Pinned Connectors of each SC to MTPConnector jumper to the system MTP Pinless Connectors.Record the measurement for fiber one. Disconnect the firstSC connectors of each MTP Connector to SC cable assem-bly and reconnect with the second SC connectors. Recordthe measurement for fiber two, then repeat for all 12 fibers.

For additional information, please refer to Corning Cable Systems ApplicationsEngineering Note AEN 78 – Field Test Procedure for Measuring Optical PowerLoss of MTP (Pinless) Connector Links.

Application GuidelinesTesting the attenuation of each segment from patch panelto patch panel allows the loss of virtually any path to bedetermined by adding the loss of the segments involved.This testing will ensure predicted system performance,document the system as built and allow routine mainte-nance checks.

The current TIA/EIA 568 Rev. C standard recommendsend-to-end attenuation tests on both specified wavelengthsfor every connectorized fiber in the backbone and tests atone wavelength in horizontal segments. Based on currentmultimode deployments where the vast majority of applica-tions use 850 nm transceivers, Corning Cable Systemsrecommends determining if 1300 nm testing is necessaryin the backbone by reviewing potential future protocols

for the system. If 1300 nm testing is unnecessary, considertesting at 850 nm only. Single-mode fiber should still betested in one direction at both 1310 nm and 1550 nm.

Acceptable link attenuation or system budget is dependenton the backbone length, the number of splices, and thenumber of connector pairs. Unless otherwise specified, max-imum acceptable fiber attenuation values can be determinedfrom the cable data sheet or the manufacturer’s specifica-tions. The attenuation value (dB/km) multiplied by length(km), will give you the maximum fiber attenuation (dB).Furthermore, if the link contains splices or connector pairs,add 0.3 dB per splice point and 0.75 dB per connector pairper TIA/EIA-568-C.3.

For example, a system that has 1.6 km of fiber, two connec-tor pairs and two splices. If the fiber in the cable is 50/125µm, the maximum fiber loss is 1.6 km multiplied by 3.5dB/km @ 850 nm and 1.5 dB/km @ 1300 nm for valuesof 4.2 dB @ 850 nm and 1.8 dB @ 1300 nm. With a totalconnector loss of 1.5 dB and a total splice loss of 0.6 dB,the budget will be 7.7 dB @ 850 nm and 4.5 dB @ 1300 nm.

OTDR TestingEnd-to-end attenuation testing measures the total amountof loss between two end points. To find out what causesthis loss and where it occurs in the cable system, anOptical Time Domain Reflectometer (OTDR) is needed.An OTDR can locate fiber events and measure the lossesattributable to cable, connectors, splices and/or othercomponents. The graphical display of loss over a cable’sentire length provides the most revealing analysis anddocumentation available on a cable link, commonlyreferred to as its signature trace.

SC Connectors

MTP Pinned Connector


Link Under Test

Blue Leg Blue Leg

Aqua LegAqua Leg

SC Connectors

Source Meter1.5 dBm

OTS-600

1

4

8

0

7 9

5 6

2 3

OTS-600

1

4

8

0

7 9

5 6

2 3

Figure 12.10Testing the MTP Connector Link Starting with the Blue Fiber| Drawing ZA-3596


Because of the OTDRs ability to provide detailed analysisof individual installed components with access to onlyone end of the fiber, it is the most versatile installationand troubleshooting tool that can be used in a varietyof scenarios:

• Cable Acceptance – evaluates the integrity, overalllength and fiber attenuation in dB/km for cables beforeand after installation. This is useful for checking a cableagainst specification, uncovering point defects due tohandling during transport or installation, and effectivelymeasuring unterminated fibers.

• OTDR Signature Trace Documentation – providesuseful documentation for cable system acceptance,network planning, and maintenance as the “as-built”fiber blueprint.

• Connector and Splice Loss – measures and documentsfield-installed connectors and midspan mechanical orfusion splices. This allows the installer to determinewhether a splice or connector is acceptable or needs tobe reworked.

• Troubleshooting – provides both (a) a benchmark ofinitial system performance for comparisons over timeand (b) a powerful tool for identifying and locating cableproblems or breaks by accessing only one end of thecable. Fiber discontinuities and localized losses are clearlyvisible when compared to original signature traces.

Background and Trace InterpolationAn OTDR works a lot like radar, sending pulses of laserlight out through the fiber and then precisely measuringthe level and time delay of the reflected pulses as theyreturn. The OTDR presents this as loss and distanceinformation in graphical format, providing a detailedoverview of the entire cable length at once. Figure 12.11and Figure 12.12 shows the OV-1000 OTDR and asample OTDR signature trace.

• The OTDR plots distance in meters or feet on the hori-zontal scale and relative loss in dB on the vertical scale.The overall trace declines from left to right, indicatingthat the light is being attenuated by the fiber, connectorsand splices as it travels down the length of the cable.Linear sections represent continuous spans of cable.

• Slopes indicate distributed loss over a section of fiber(steeper slopes indicate higher fiber loss in dB/km).

• Vertical drops represent point losses at connectors,splices and faults. The magnitude of the droprepresents loss in dB.

• Spikes or humps indicate reflective events such as con-nectors or mechanical splices where the continuity ofthe glass is interrupted. The final spike on the traceindicates the end of the fiber.

• Test fiber boxes are required to mitigate the effects ofOTDR high-powered launches which may saturate theOTDR receiver. This generates an inaccurate trace forthe first several meters of the tested system. A minimumlength of 100 m for multimode systems and 300 m forsingle-mode systems is required. Test fiber boxes arethe same fiber core diameter as the system lengthbeing tested.

To allow measurement of the connector loss at the opticalpatch panel, a test fiber box is used to connect between theOTDR output and the interconnect hardware.

Figure 12.11OV-1000 OTDR | Photo LAN731

Figure 12.12Sample OTDR Signature Trace | Drawing ZA-3659


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

The OTDR user can place markers and cursors on thetrace to make measurements more easily and reliable.Some OTDRs and multifunction testers provide a functionthat automatically configures the OTDR, performs asignature trace and measures the position and loss of eachevent in the cable system. Figure 12.13 shows an exampleof an OTDR table summarizing the event data gathered insuch a function.

For further information on OTDR measurements, refer tothese Corning Cable Systems Application Engineering Notes:AE Note 003, Unidirectional Single-mode MeasurementsAE Note 007, GainersAE Note 033, OTDR Return Loss MeasurementAE Note 036, Optical Fiber Fault Location ProcedureAE Note 050, “Ghost” Reflections on the OTDR

Connector and Splice Loss MeasurementFor all cable segments, Corning Cable Systems recommendsOTDR measurement of each field-installed connectorand each mechanical or fusion splice at one wavelengthto ensure they meet acceptable loss values certified bythe installer. Loss values from some manufacturers’ splicemachines can be substituted for OTDR measurementsprovided they employ either a LID-SYSTEM™ Unit orLPAS system to obtain splice loss values. Again, unlessspecified otherwise, acceptable losses are ≤ 0.75 dB permated connection and ≤ 0.3 dB per splice for multimodeand single-mode. To measure a near-end connector loss,a test fiber box of sufficient length (typically ≥ 100 m formultimode or ≥ 300 m for single-mode) is used to connectbetween the OTDR and patch panel and has a cable spanof ≥ 75 m following the connector.

User Tip: Use of a test fiber box also allows simultaneous OTDR testing of a link’ssignature trace and near-end connector loss. These test results can be documentedtogether on OTDRs that have event tables.

Test Equipment: OTDR AnalysisA variety of units incorporating the OTDR concept areavailable. More useful OTDRs include:

• Dual 850/1300 nm multimode and 1310/1550 nmsingle-mode operation in the same unit; whether or notsingle-mode is used today, the unit should be upgradeableto meet future requirements

• Portable, battery-powered operation

• An internal flash, USB port and hard drivefor trace storage

• A companion PC software package for analysis,comparison and printing of saved traces

• Integrate power meter and visual fault locator functionsinto the same unit, maximizing the unit’s utility andcost-effectiveness.

• Combining a multi-tester with an optical source allowsend-to-end attenuation test results to be stored in a fileand associated with their respective OTDR traces.

As the number of fibers and cable systems increases, thesefiber management and documentation features save timeand effort.

Bandwidth and DispersionBandwidth and dispersion measure characteristics ofthe information of the information carrying capacity offiber. Fiber optic cable can be specified for various gradesof bandwidth or dispersion performance. Actual systembandwidth or dispersion is a function of the fiber quality,length and transmitter characteristics. It is common practiceto specify cable bandwidth and dispersion performanceto meet the requirements of TIA/EIA-568-C as discussedin Chapter 3, ensuring compatibility with transmissionelectronics without field testing. The fiber manufacturer’sbandwidth or dispersion performance should be documentedon a specification sheet and saved for future reconfigurationsand upgrades.

Figure 12.13Sample OTDR Table | Drawing ZA-2892


Summary of Cable Systems TestingRecommendationsBased on the characteristics of backbone and horizontalcable segments, Table 12.2 summarizes both multimodeand single-mode test recommendations for all connectorizedfibers. If any fibers are left unterminated, Corning CableSystems recommends performing an OTDR inspectionfor spans longer than 75 m.

DocumentationDocumentation plays a vital role in the long-term successof any cabling system with regard to system reconfiguration,upgrades and maintenance. End-to-end test resultsestablish the initial integrity and performance of a system.Documents of work performed on the fiber plant can beused for liability protection in the event that multiplevendors are involved. Equally important, these recordsestablish “as-built drawings” and can be compared tocurrent conditions when troubleshooting.

Careful planning and accessible documentation also helpto avoid costly retesting or cable plant replacement whensubsequent upgrades or reconfigurations are undertaken.Following the requirements stated in TIA/EIA-606,Corning Cable Systems recommends maintainingaccessible documentation of the following test resultsand cable records.

Test Results• End-to-end attenuation data

• OTDR signature traces

• Certificate of compliance for connector and splice loss

Cable Records• Cable specifications

• Cable route diagrama. Fiber routing and location informationb. Fiber connectivity informationc. Splice point locationsd. Patch panel locationse. Cable lengthsf. Cable part numbers

Data Center Segment

Backbone Horizontal EquipmentTest Method Cabling Cabling (Multimode) Required

End-to-End Dual wavelength insertion loss 850 or 1300 nm • Optical MeterAttenuation • Multimode: 850 and 1300 nm (multimode) • Optical Source(s)(Required) • Single-mode: 1310 and 1550 nm • Mandrel

• Two Test Jumpers• One Adapter

OTDR Test OTDR inspection of each Fiber > 75m Troubleshooting • OTDR(Optional Only as required for links • Test Fiber Boxfor Inside Plant) • Multimode: 850 and 1300 nm exceeding the

• Single-mode: 1310 and 1550 nm budgets dB limit

Dual wavelength or bi-directionaltesting as required

OTDR inspection of each field-terminated connector and eachsplice at one wavelength:

Note: Simultaneous testing of a fiber’ssignature trace (above) and near-endconnector loss can save test timeand documentation.

TABLE 12.2: TIA-568-C.1 System Testing Recommendations


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Note: For Corning Cable Systems guidelines, see page 58.

Maintenance and TroubleshootingBecause of the quality and importance of informationtransmitted over fiber optic systems, ongoing service iscritical. A properly installed and tested system requiresminimal routine maintenance. Ensuring proper connectorcare and cleanliness and checking the routing and protectionof system jumpers are simple safeguards that are centralto preventing possible service interruptions.

In the case of system error or failure, troubleshootingand service restoration can be performed quickly and easily.There are three key components required for efficienttroubleshooting:

• Documentation – Initial test results and cable recordsare essential to effective maintenance and troubleshoot-ing. Contrasting current test results with the originaldocumentation quickly and clearly identifies changesand potential trouble spots.

• Test Equipment – Using a simple power meter andinitial attenuation test results to isolate faults willeliminate unnecessary service calls and minimizedowntime. Faulty patch cords can be replaced. If thefault lies within the cable plant, an OTDR can be usedto pinpoint its exact location.

• Troubleshooting Plan – A simple but effective flowchart or procedure can be used to quickly isolate a faultto either a network transmitter, receiver, patch cordor cable segment. The first step requires only a powermeter, test jumper and the “as-built” documentation.

Troubleshooting Process FlowTroubleshooting can be very difficult but can be made easywith a defined process. Using a basic fiber system, Tx→plant→ Rx, one can develop a basic process for testing.First, the received power level is measured and comparedto the receiver sensitivity specification using a source andmeter test such as Corning Cable Systems’ OTS-600. If thereceived power is normal, the receiving electronics shouldbe diagnosed to identify the problem. If, on the other hand,the received power level is low, the transmitter outputpower should be measured next. A low transmitter outputindicates a problem with the transmitter output or electron-ics. In these cases, follow the procedure in diagnosing theelectronics or call the appropriate vendor for assistance.

If the transmitter output is normal and the received poweris low, excessive loss is occurring in the cable plant. A powermeter with a test jumper is then used to confirm whether ornot there is a problem with the system jumper. If the systemjumpers have acceptable loss, then the fault probably lieswithin the terminated cable plant itself. Losses in the cableplant are most often caused by damaged connectors and cutor damaged cable.

Once a problem is isolated to the cable plant, a high-resolu-tion OTDR such as Corning Cable Systems’ OV-1000 isused to locate the fault. A comparison of the original signa-ture trace to the current OTDR trace can easily identify andlocate a fault or break. If the fault is determined to be nearan end-point, a visual fault locator such as Corning CableSystems’ VFL-350 can be used to pinpoint a problem withinsplice trays, connecting hardware and patch cords.


An important step in the documentation process is properlabeling of all the data center infrastructure components.Every component of the telecommunications infrastructureshould be labeled in an independent manner. For purposesof tracking the fiber and documentation, the most importantthings to keep in mind with the labeling system are the localand remote site terminations defined by the location of thetelecommunications building, rooms, rack or cabinet, fiberpanels and fiber or port ID and the fiber itself.

Choosing a Labeling MethodThe ANSI/TIA-606-A-1 standard specifies administrationfor a generic telecommunications cabling system thatwill support a multi-product, multi-vendor environment.It provides a uniform administration approach that is inde-pendent of applications, which may change several timesthroughout the life of the telecommunications infrastruc-ture. It establishes guidelines for owners, end users, manu-facturers, consultants, contractors, designers, installers andfacilities administrators involved in the administration ofthe telecommunications infrastructure.

Labeling Racks and CabinetsWith today’s data centers, finding the right patch paneland port starts with quickly finding the rack or cabinet thathouses the patch panel. Some data center administrators havecreated their own system for identifying cabinets or racks ina data center, but TIA-606-A-1 is meant to help streamlinethe process, promote a consistent and reliable methodologyand make it easier on the data center administrator. Creatingrack/cabinet identifiers in the data center is accomplishedby using X and Y coordinates that relate to floor tiles in araised-floor system or to the number or rows and cabinetsin a data center floor plan. The “X” coordinate is an alphacharacter and the “Y” numeric, resulting in a unique identifi-er for each rack and cabinet. TIA-606-A-1 specifies that therack/cabinet identifier label shall be placed at the top andbottom on both the front and rear of each rack or cabinet.

Each telecommunications space also has a unique identifierwhere “F” is a numeric character that identifies the floor ofthe building and “S” is an alpha character that defines thespace. The XY cabinet identifier follows the FS identifier(FS.XY), creating a specific location for racks and cabinetsthat can be applied to any space.

Chapter Thirteen:Labeling

AA AB AC AD AE AF AG AH AI AJ AK AL AM AN AO AP AQ AR AS AT01

02

03

04

05

06

07

08

09

10

11

12

Sample Cabinet AG04





Front - EndLayer Zone


Back - EndLayer Zone

StorageZone

24F 24F 24F 24F 24F 24F 24F 24F

3 x 36F 3 x 36F


(MDA)

Y-axis

X-axis

Figure 13.1“Grid Coordinate” System for Data Center Equipment | Drawing ZA-3660

Chapter Thirteen: Labeling | LAN-1160-EN | Page 63

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Labeling Patch PanelsTIA 606-A also contains additional identifiers for patchpanels and ports or fiber. Most commonly, patch panelidentifiers are numeric and designate the top left cornerof the patch panel location starting from the bottom ofthe frame or cabinet and are specified in rack units.

Individual panels, modules and port or fiber identificationwith the patch panel are also accounted for in documentingthe local and remote fiber termination location. Fiberpanels should be clearly labeled with port or fiber ID.Documentation should clearly identify what fiber strandsare connected to which bulkhead and typically specify therange of ports or fibers the trunk is servicing locally andremotely.

The documented code that is printed and recorded on thelabel can be constructed by combining the pertinent identi-fiers for each hardware component for the local and remotelocation of the fiber terminations by properly identifyingthe elements of the infrastructure. This code can be usedto track each component of the infrastructure.

Labeling FiberFiber termination identification is equally important in map-ping the data center network. Individual fibers (such asjumpers) must be clearly labeled to identify local and remotelocation. Typically a single jumper will contain two labels oneach end of the fiber near the termination point identifyingthe specific local ID and the remote ID the jumper is patch-ing to. It is important to follow suit with previous definedcode indicating rack or cabinet ID, patch panel ID and morespecifically the individual port or fiber termination location.This allows users to easily trace jumpers from one locationwithin the network to the next. Most common identificationmethods for individual fiber ID employ the use of flag orwraparound preprinted labels. Documentation should clearlyidentify individual fiber strands of the cable or jumper.

Numeric identifiers for cables and cable strands can be usedsolely to differentiate them from other cables sharing theirsame characteristics.

ExampleThe code shown on page 65 provides an analogous solutionin accordance with TIA-606-A-1 for mapping to beemployed with Corning Cable Systems Pretium EDGE™

Solutions hardware.

Detailed and accurate record keeping enables users tologically “map” fiber terminations within the data centerfrom local equipment to remote equipment. It is suggestedthat users employ labeling guidelines demonstrated inEIA/TIA-606 for mapping the network. Guidelines belowprovide an analogous solution in accordance with TIA-606for labeling to be employed with Pretium EDGESolutions hardware. Additions in the suggested codingare accounted for to identify chassis trays and modules.

Hardware Labeling• Hardware components come pre-labeled for

identification and promote a consistent labeling scheme.• User must provide their own label maker and media for

the frame or cabinet and hardware.• Labeling is best supported with use of adhesive-backed

label makers with media up to one-half inch in height.• Front door of the chassis has a locating crop mark

suggesting a consistent location for the printed ID labelthat identifies the location of the chassis within the frameor cabinet.

• Chassis trays are pre-labeled 01 to 12 from the bottomof the chassis to the top.

• Module positions within each chassis tray are identifiedby alpha characters A through D from left to right.

• Modules or MTP® panels come assembled with adaptersand silk screened with fiber and/or port ID.

• Chassis comes equipped with a label card that is easilyremovable from the inside of the front door and requiresno additional fastening to remain in place. This labelmay be written on but use of a label maker is best.

• Label card supports adhesive-backed label media andprinted labels may be easily adhered or removed formoves, adds and changes.

Cable/Jumper Labeling• Labeling of cable and jumpers is equally important

and can be completed following same suggestedcoding system shown on page 65 and cross reference torecommended guidelines of EIA/TIA-606. A preprintedcode on flag/wraparound labels is suggested and may bepurchased from many media suppliers.


Hardware labeling involves a 5-step process that identifiesthe local and remote site. Use the suggested code in Figure13.2 and the following steps to map your location.

STEP 1: Frame or cabinet locationSTEP 2: Chassis locationSTEP 3: Tray locationSTEP 4: Module location with fiber or port rangeSTEP 5: Documentation

STEP 1 – Frame or Cabinet Location (Figure 13.3)Identify location of frame or cabinet within the floor spacegrid coordinate system. Preprint labels and adhere to thefront and back of the frame or cabinet at the top and bottom.

f1 s1 x1 T-t1 m1pn1y1-CODE

STEP 1 - Frame / Cabinet locationf1 = "Optional" Floor of the buildings1 = "Optional" Telecom Space IDFloor space grid coordinate location of Frame or Cabinet defined by:x1 = Two Alpha indicating rowy1 = Two Numeric indicating position

LOCAL ID

- TO:z1a1 f2 s2 x2 T-t2 m2pn2y2- - :z2a2

12 N AJ T-06 A1-1204-Ex: - TO: 21-1A80-TKBN2182 09- - :40

REMOTE ID

STEP 2 - Chassis locationz1 = "Optional" identifies Front or Back of the Frame or CabinetUse "F" for Front or "B" for Backa1 = Location of top / left corner of chassis within Frame or CabinetSpecified in Rack Units

STEP 3 - Tray locationT-t1t1 = Tray location within chassis

STEP 4 - Module location with Fiber or Port Rangem1 = Module location within Tray (A, B, C, or D)pn1 = Fiber or Port range within Module

STEP 5 - DocumentationPrinted Label records Remotetermination of Trunk Cable

Figure 13.2Suggested Code For Labeling | Drawing ZA-3661

Figure 13.3Step 1: Frame or Cabinet Location | Drawing ZA-3662


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

STEP 2 – Chassis Location (Figure 13.4)Identify location within the frame or cabinet (in rack unitsfrom the bottom) by locating the top/left corner of thechassis. Print two labels and adhere one to the front doorusing the crop mark for alignment. Adhere the secondlabel to label card on the inside of the door.

STEP 3 – Tray Location (Figure 13.5)Identify location of tray within the chassis.Trays come pre-labeled 01 to 12 from the bottomto the top of the chassis.

STEP 4 – Module Location with Fiberor Port Range (Figure 13.6)Each tray comes pre-labeled A through D to identifythe module position within the tray.

STEP 5 – Documentation (Figure 13.7)Identify remote location and determine code to be printedfollowing Figure 13.7. It should only be necessary to printthe remote location. Adhesive labels may be affixed to thelabel card on the inside of the front door.

PRINTED LABEL RECORDSREMOTE TERMINATION OF

TRUNK CABLE

12N-AJ04-28

TO 12N-BK09-40 T-08:A1-12

12N-AJ04-28

TO 12N-BK09-40 T-08:A1-12

Figure 13.7Step 5: Documentation | Drawing ZA-3666

Figure 13.4Step 2: Chassis Location | Drawing ZA-3663

12N - AJ04

12N - AJ04

06

Figure 13.5Step 3: Tray Location | Drawing ZA-3664

FIBER ID

LC MODULE

PORT ID

MODULE LOCATIONWITHIN A TRAY

FIBER or PORT RANGEWITHIN MODULE

AB

AB

CD

21 1

243

Figure 13.6Step 4: Module Location | Drawing ZA-3665


ABFAir-blown fiber. An alternate fiber provisioning schemewhich requires pre-provisioning plastic tubes to all possibleservice locations. Fiber is later selectively installed to serv-ice locations. ABF is not compliant to TIA-568 standardsand is not accommodated by typical building constructionpractices.

Acceptance ConeAn imaginary cone that defines the angle with which anoptical fiber will accept incoming light.

Access JumperA length of fiber placed between the Optical Time DomainReflectometer (OTDR) and an event along a fiber that is tobe measured. This allows the user to see fiber on both sidesof the event so that its loss can be estimated. Length mustbe significantly greater than the OTDR attenuation deadzone. A mechanical media termination device designed toalign and join fiber optic connectors; often referred to as acoupling, bulkhead or interconnect sleeve.

AdapterA mechanical media termination device designed to alignand join optical fiber connectors; it is often referred to as acoupling or interconnect sleeve.

AHJAuthority having jurisdiction. The organization, officeand/or individual responsible for approving equipment,an installation, or a procedure. Note: the phrase “authorityhaving jurisdiction” is used in a broad manner sincejurisdictions and approval agencies vary as do theirresponsibilities.

ALTOS® CableCorning Cable Systems’ stranded loose-tube cable in whichbuffer tubes contain two or more fibers and which usesinnovative waterblocking technology for craft-friendliness.

AMAmplitude modulation. An analog signal with a constantfrequency and varying amplitude.

Anaerobic-Cure ConnectorA field-installable connector with a polymer epoxy thathardens when combined with an activating agent.

AnalogA communications format that uses continuous physicalvariables such as voltage amplitude or frequency variationsto transmit information.

ANSIAmerican National Standard Institute

Aramid YarnStrength elements that contribute cable tensile strength,support and additional protection of the optical fiberbundles.

Arbitrated LoopFibre Channel topology in which devices are connectedin a loop; a token is used to control access.

ArmorAdditional protective element beneath the cable outerjacket used to provide protection against severe outdoorenvironments and gnawing rodents. It is usually made ofplastic-coated steel and it may be corrugated for flexibility.

As-Built TestTest performed after all installations (cable placement,splicing, connectorization) have been completed, to showthe system performs to specifications; usually comprisedof OTDR and end-to-end attenuation tests.

ATM (Asynchronous Transfer Mode)A network communications protocol standard witha digital transmission switching format; designed forscalable bandwidth and multimedia voice, data andvideo transmission.

AttenuationThe decrease in magnitude of signal power transmittedbetween points; a term used for expressing the total lossof an optical system, normally measured in decibels (dB)at a specific wavelength.

GlossaryThe following are terms used within this guide. These terms are defined within the context of the optical fiber industry.

Glossary | LAN-1160-EN | Page 67

INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Information and Tools 5SECTION

Attenuation CoefficientThe rate of optical power loss with respect to distancealong the fiber, usually measured in decibels per kilometer(dB/km) at a specific wavelength; the lower the number,the better the fiber’s attenuation. Typical multimode wave-lengths are 850 and 1300 nanometers (nm); single-modewavelengths are 1310 and 1550 nm. Note: When specifyingattenuation, it is important to note the value is maximum.

Backbone (Data Center)Provides interconnection between the main distributionarea, the horizontal distribution area and entrance facilities.

Backbone Cabling (LAN)The portion of premises telecommunications cabling thatprovides connections between telecommunications closets,equipment rooms and entrance facilities. The backbonecabling consists of the transmission media (optical fibercable or copper twisted-pair), main and intermediatecross-connects, and terminations for the horizontal cross-connect, equipment rooms, and entrance facilities. Thebackbone cabling can further be classified as interbuildingbackbone (cabling between buildings) or intrabuildingbackbone (cabling within a building).

BackscatterThe portion of light that is scattered by the structure ofthe glass and travels back toward the source. The OTDRuses this scattered light to make measurements.

BandwidthMeasure of the information-carrying capacity of an opticalfiber usually measured in MHz•km at a specific wave-length. The higher the bandwidth, the better the fiber.Note: This term is often used to specify the normalized modal bandwidth of a

multimode fiber.

Base-2 Cabling SystemsData center backbone and horizontal cables that areterminated on each end with a 2-fiber duplex connector.

Base-12 Cabling SystemsData center backbone and horizontal cables that areterminated on each end with a 12-fiber MPO connector.

Base-24 Cabling SystemsData center backbone and horizontal cables that areterminated on each end with a 24-fiber MPO connector.

Bend-Radius (Fiber)Radius a fiber can bend before the risk of breakage orincrease in attenuation. See Cable Bend-Radius.

BroadbandDenotes transmission facilities capable of handling awide range of frequencies simultaneously, thus permittingmultiple channels in communications systems. It isnormally associated with CATV systems.

Buffering(1) A protective material extruded directly or aroundthe coated fiber to protect it from the environment(also known as tight-buffered); (2) extruding a tube aroundcolored fiber to allow isolation of the fiber from stressesin the cable (also known as buffer tubes).

Buffer TubesExtruded cylindrical tubes covering optical fiber(s) used forprotection and isolation. See Loose Tube.

BulkheadSee Adapter.

BundleMany individual fibers contained within a single jacket orbuffer tube. Also, a group of buffered fibers distinguishedin some fashion from another group in the same cable core.

BWBandwidth

ByteA sequence of 8 bits.

CabinetA physical enclosure for rack-mountable equipment.Cable, optical fibers and other material(s) assembledto provide mechanical and environmental protectionfor the fibers.

Cable AssemblyOptical fiber cable with connectors installed on one orboth ends. Cable assemblies are generally used for inter-connection of optical fiber cable systems and opto-elec-tronic equipment. If connectors are attached to only oneend of a cable, it is known as a pigtail. If connectors areattached to both ends of a low-fiber-count cable, it isknown as a jumper or patch cord.


Cable Bend-RadiusCable bend-radius during installation is the smallest radiusbend for a cable experiencing a tensile load. Cable bend-radius installed is the smallest diameter bend for a cablethat is under no tensile load.

CamSplice™ Mechanical SpliceCorning Cable Systems’ non-adhesive mechanical splice.

Carrier Sense Multiple Access/CollisionDetection (SMA/CD)This is the communication scheme used in a sharedEthernet network.

CascadeAn architecture in which switches are daisy-chainedtogether. Frames are passed from switch to switch untilthe port for the destination device is reached.

CATVCommunity access television

CCHCloset connector housing

CCSCloset connector and splice housing

CDFCloset distribution frame

Central MemberThe center component of a stranded loose tube cable. Itserves as an anti-buckling element to resist temperature-induced stresses. The central member material is steel,fiberglass or glass-reinforced plastic (GRP).

Centralized CablingA cabling topology used with centralized electronics con-necting the optical horizontal cabling with intrabuildingbackbone cabling passively in the telecommunicationscloset or main cross-connect.

ChannelA dedicated path between two devices characterized byvery high data rates and very low overhead; it is typicallyhardware intensive and addresses system data as part ofthe “setup” information.

Chromatic DispersionSignal dispersion caused by light traveling at multiplewavelengths which arrive at the detector at different times.

CJPCloset jumper-management panel

CladdingThe material surrounding the core of an optical fiber.The cladding must have a lower index of refraction tokeep the light in the core.

Class of ServiceThe four classes include connection oriented, connection-less, datagram and fractional bandwidth services.

CoatingA material applied to a fiber during the manufacturingprocess to protect it from the environment and handling.

Coaxial CableA central conductor surrounded by an insulator, which inturn is surrounded by a tubular outer conductor, which iscovered by more insulant; also called coax cable.

CollisionThe result when two users attempt to send data simultane-ously on a shared media network. Data is corrupted andboth devices must retransmit their information.

Composite CableA cable containing both fiber and copper media.

Computer RoomAn architectural space to accommodate dataprocessing equipment.

ConduitPipe or tubing through which cables can be pulledor housed.

Connecting HardwareA device used to terminate an optical fiber cable withconnectors and adapters providing an administrationpoint for cross-connecting between cabling segmentsor interconnecting to electronic equipment.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

ConnectorA mechanical device used to align and join two fiberstogether to provide a means for attaching to andde-coupling from a transmitter, receiver or another fiber.Commonly used connectors include the MT-RJ, SC,ST® Compatible and LC connectors.

Connector ModuleA connector panel with a pre-installed cable assembly(or assemblies) on the back plane, which can be spliced tobackbone cable fibers (designed for use with patch panels).

Connector PanelA panel insert designed for use with patch panel housings.Connector panels often contain pre-installed adapters.

CoreThe central region of an optical fiber through which lightis transmitted.

CouplingSee Adapter.

Cross-ConnectIncoming and outgoing fibers terminated in adapter sleevesor the backplane of the patch panel. Single-fiber jumpers,which are installed on the front plane, complete the circuits.

Cross-Connect SwitchA fabric switch that connects only to other switches (I/O).

CSHCloset splice housing

Composite Second-Order Beat (CSO)A clustering of second-order beats 1.25 MHz abovethe visual carriers in CATV systems.

Composite Triple Beat (CTB)A clustering of third-order distortion products aroundthe visual carriers in CATV systems.

Corning® ClearCurve® Multimode Optical FiberThe world's first laser-optimized multimode fiber towithstand tight bends at or below 10 mm radius withsubstantially less signal loss than traditional multimodefiber. This new fiber allows designers, installers andoperators of enterprise networks - including local areanetworks, data centers and industrial networks - to deployoptical fiber in more places by delivering all of the band-width benefits of optical fiber in a package that is easierto handle and install than copper.

Cut-Off WavelengthThe wavelength below which a single-mode fiber willsupport more than one mode of light.

CWDMCoarse wavelength division multiplexing

Data CenterA building or portion of a building whose primary functionis to house a computer room and its support areas.

dBSee Decibel.

Dead ZoneAttenuation dead zone is the distance after a reflectiveevent at which the trace line has returned to within 0.5 dBof the actual backscatter line. It is caused by the laser pulsereflecting as it passes through the connection or event.

DecibelThe unit for measuring the relative strength of light signalsexpressed as dB. It is equal to one-tenth the commonlogarithm of the ratio of the two power levels. It is expressedin dBm when a power level is compared to 1 milliwatt.

Demarcation PointA point where the operational control or ownershipchanges.

DielectricNon-metallic electrically non-conductive. Glass fibersare considered dielectric. A dielectric cable contains nometallic components.

DigitalA data format that uses discrete physical levels totransmit information.

DispersionThe broadening of light pulses along a length of the fiber.Two major types are (1) modal dispersion caused by differ-ent optical path lengths in a multimode fiber; (2) chromat-ic dispersion which is the sum of material dispersion andwaveguide dispersion in single-mode fiber. Material dis-persion is pulse spread caused by different index of refrac-tion for light of various wavelengths in a waveguide mate-rial. Waveguide dispersion is caused by light traveling atdifferent speeds in the core and cladding of single-modefibers with the spreading of a light pulse as it travels downa fiber. The higher the dispersion, the lower the maximumtransmission frequency.


Dispersion-Shifted FiberSingle-mode fiber that has a zero dispersion wavelengthin the 1500 nm region.

Distributed Feedback Laser (DFB)Edge-emitting laser typically used for 1310 nm/1550 nmoperation.

DMDDifferential modal delay

DocumentationThe methodical recording of test and physical data for afiber system, including OTDR traces, end-to-end losses,connector and splice losses, route diagrams, meter/footmarks such that a complete record is produced of theactive condition of the completed system.

DSPDigital signal processing

DTEData terminal equipment

Duplex ConnectorTwo connectors mechanically joined side by side;terminating two separate strands of fiber.

EDAEquipment distribution area. The computer room spaceoccupied by equipment racks or cabinets.

EDCEnvironmental Distribution Center

EDGESee Pretium EDGE™

Effective Modal Bandwidth (EMB)The system modal bandwidth observed in a link for aspecific fiber with a specific source.

Effective Modal Bandwidth, Calculated (EMBc)It predicts source fiber performance by integrating thefundamental properties of light sources with the multi-mode fiber’s modal structure ensuring that the effectivemodal bandwidth (EMB) of a fiber will meet the 10 Gb/srequirement of 2000 MHz•km with any conforming laser.

Electromagnetic Interference (EMI)Radiated or conducted electromagnetic energy that hasan undesirable effect on electronic equipment or signaltransmissions.

End-to-End TestMeasurement of optical power loss using a source andmeter which transmits into one end of the fiber andreceives at the other end; typically from one patch panelto another.

Entrance FacilityAn entrance to a building for both public and privatenetwork service cables including the entrance point atthe building wall and continuing to the equipment roomor space.

Entrance Room (ER)A space in which the joining of interbuilding orintrabuilding telecommunications backbone facilitiestakes place.

Equipment RoomA centralized space for telecommunications equipmentthat serves the occupants of a building. An equipmentroom is considered distinct from a telecommunicationscloset because of the nature or complexity of theequipment.

EthernetAn IEEE network protocol standard for a 10 Mb/s localarea network. The IEEE 802.3 standard defines thevarious requirements and speeds of Ethernet that include10 Mb/s, 100 Mb/s, 1000 Mb/s (1 Gb/s) and 10 Gb/sEthernet. Also see Fast Ethernet, Gigabit Ethernet and10 Gigabit Ethernet.

EventAny component, such as connectors, splices, faults etc.that is displayed on an OTDR trace.

Event SearchAn OTDR’s ability to use an algorithm to search,automatically, for all events in the cable, reporting theirlocation and loss.

FabricTopology using switches to connect one or multipledevices to other devices that are part of the network.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Fan-OutCorning Cable Systems’ tight-buffered breakout-stylemulti-fiber cable designed for ease of connectorizationand rugged applications for interbuilding or intrabuildingrequirements.

Fast EthernetEthernet at 100 Mb/s transmission rate. This is defined bythe IEEE 802.3 standard.

FCCFederal Communications Commission

FCoETransmission method in which the Fibre Channel frameis encapsulated into an Ethernet frame at the server.

FerruleA mechanical component, generally a rigid ceramic tube,used to protect and align a fiber in a connector.

FiberThin filament of glass; an optical waveguide consisting of acore and a cladding that is capable of carrying informationin the form of light.

Fiber Bend-RadiusMinimum radius a fiber can bend without experiencinga reduction in optical fiber reliability.

Fiber Distributed Data Interface (FDDI)A standard for a 100 Mb/s fiber optic local area network.

Fiber IdentifierA device that bends a fiber (slightly) so that enough lightleaks out that a detection can determine the presence oftraffic and its direction, as well as recognize the presenceof a test tone (usually 2 kHz).

Fiber OpticsLight transmission through optical fibers for communica-tion or signaling.

Fibre ChannelConnecting protocol commonly used in data centers tolink servers to storage arrays. Fibre Channel mandatesreliable delivery of data. Common data rates are 1 Gb/s,2 Gb/s, 4 Gb/s, 8Gb/s and 10 Gb/s.

Field-Installable ConnectorAn optical connector that can be assembled in the field(at the job site) and installed by hand.

FMFrequency modulation

FOTPFiber optic test procedures; defined in TIA/EIAPublication Series 455.

FPFabry perot (laser)

FrameThe smallest subset of data; frames make up sequences.

Fresnel Reflection LossesReflection losses that are incurred at the input and outputof optical fibers due to the differences in refraction indexbetween the core glass and immersion medium.

FTTxFiber to the x. A growing practice of provisioningindividual subscribers with 100 percent optical fiber fromthe POP to the premises. The parity cost of fiber relativeto copper plant and its extraordinary bandwidth advantagehave made FTTx economically attractive in manyapplications.

Full-DuplexCapable of transmitting and receiving over the samechannel simultaneously. In pure digital networks, this isachieved with two optical fibers.

Functional LevelsThe model (consisting of five levels) that defines FibreChannel operation. These levels include the physical media,encoding scheme, frame layout and services mapping.

FusingThe actual operation of joining fibers together by fusionor by melting.

Fusion SpliceA permanent joint produced by the application of localizedheat sufficient to fuse the ends of two optical fibers,forming a continuous single-light path.

FZBFiber zone box


GainerA splice loss measurement in which the trace appears to goup (more power) and there appears to be a gain, instead ofa loss; typical in cases where fiber of differing manufactur-ers is spliced together. Testing from the opposite directionusually produces a corresponding loss equal to the powergain measured from the other direction.

GbESee Gigabit Ethernet.

GhostAn “echo” caused by highly reflective components (connec-tors) in which light is reflected back from the connection,strikes another connection, which reflects it back out intothe fiber, only to be reflected back to the OTDR again.

Gigabit EthernetA 1000 Mb/s transmission rate. This is defined by theIEEE 802.3 standard.

Gigahertz (GHz)A unit of frequency that is equal to one billion cyclesper second.

Graded-IndexMultimode fiber design in which the refractive indexof the core is lower toward the outside of the fiber coreand higher toward the center of the core, thus providinghigher bandwidth capabilities.

Half-DuplexThe transmission of data in both directions, but only onedirection at a time. For example, two-way radio (push-to-talk phones) use half-duplex communications. When oneparty speaks, the other party listens.

Heat-Cure ConnectorA field-installable connector with a polymer epoxy thathardens when exposed to heat.

Horizontal CablingThat portion of the LAN that provides connectivitybetween the horizontal cross-connect and the work-areatelecommunications outlet. In the data center, thehorizontal cabling provides connectivity between themain distribution area/horizontal distribution area tothe equipment distribution area. The horizontal cablingconsists of transmission media, the outlet, the terminationsof the horizontal cables and horizontal cross-connect.

Horizontal Cross-Connect (HC)The horizontal cross-connect (HC) is where the buildingbackbone and horizontal cabling meet in the telecommuni-cations room (TR).

Horizontal Distribution Area (HDA)A space in a computer room where a horizontalcross-connect is located.

HousingAn enclosure, usually metallic, for splicing or termination.

Hybrid CableA fiber optic cable containing two or more different typesof fiber, such as 62.5 µm multimode and single-mode.

ICHIndustrial connector housing

IECInternational Electrotechnical Commission

IEEEInstitute of Electrical and Electronics Engineers

Index-Matching GelA gel with an index of refraction close to that of the opticalfiber used to reduce reflections caused by refractive-indexdifferences between glass and air.

Index of RefractionThe ratio of light velocity in a vacuum to its velocity ina given transmission medium.

Insertion LossSee Loss

Intelligent Transportation System (ITS)A combination of electronics, telecommunications andinformation technology to the transportation sector forimproving safety and travel times on the transportationsystem. Intelligent transportation systems collect,store, process and distribute information relating tothe movement of people and goods.

Interbuilding BackboneThe portion of the backbone cabling between buildings.See Backbone Cabling.

Interconnect SleeveSee Adapter.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Intermediate Cross-Connect (IC)A secondary cross-connect in the backbone cablingused to mechanically terminate and administer backbonecabling between the main cross-connect and horizontalcross-connect.

International Organization for Standardization (ISO)An organization that sets international standards, foundedin 1946.

Intrabuilding BackboneThe portion of the backbone cabling within a building.See Backbone Cabling.

I/O SwitchA fabric switch that connects to both devices(input and output) and cross-connect switches.

ITUInternational Telecommunications Union

JPEGJoint picture expert group

JumperOptical fiber cable that has connectors installed on bothends and used at cross-connects and at end equipment tofacilitate patching. See Cable Assembly.

Jumper ManagementA means of providing an orderly administration of fibers.This is essential in areas of high density and should providea means of routing single-mode and multimode fibers hori-zontally, vertically, and front to back in rack installations.

Kilometer (km)One thousand meters, or approximately 3,281 ft.The kilometer is a standard unit of length measurementin fiber optics. Conversion is 1 ft = 0.3048 m.

kpsiA unit of force per area expressed in thousands of poundsper square inch; usually used as the specification for fiberproof test, e.g., 100 kpsi.

LANscape® SolutionsThe complete tip-to-tip approach to fiber cabling solutionsfor private networks that consists of a comprehensive setof integrated products, services and support to ensurea successful and efficient fiber network that will serve asa stable communications infrastructure for years to come.

LaserLight amplification by the simulated emission of radiation.A device that causes a uniform and coherent light that isvery different from an ordinary light bulb. Many lasersdeliver in an almost perfectly parallel beam (collimated)that is very pure, approaching a single wavelength. Laserlight can be focused down a tiny spot as small as a singlewavelength.

LatencyThe time delay that frames experience in traversing thenetwork, both relative to absolute time and each other.Voice and video are very sensitive to latency, whereas datagenerally is not very sensitive to latency.

Least Squares Analysis (LSA)An OTDR loss measurement made using linear regressionto determine the slope of the trace on each side of an eventand extrapolate this slope to the location of the event,determining the vertical difference at that point, whichis the loss measurement.

Lens Profile Alignment Systems (LPAS)A method of fusion splicing in which the fibers are alignedbased on the profile or the fiber. This method aligns thecladding of the fiber, not the fiber cores.

Light-Emitting Diode (LED)A display and lighting technology used in almost everyelectrical and electronic product on the market, from atiny on/off light to digital readouts, flashlights, trafficlights and perimeter lighting. LEDs are commonly usedin digital transmission sources for speeds ≤ 622 Mb/s.

LinkA telecommunications circuit between any two telecommu-nications devices, not including the equipment connector.

Local Area Network (LAN)A geographically limited communications networkintended for the local transport of voice, data and video;often referred to as a customer premises network.

Local Injection and Detection (LID)A method of fusion splicing in which a light is injectedinto the core of one fiber and sensed in the other.The fibers are aligned until the maximum amount oflight passes between them and they are fused together.

Logical vs. Physical ToplogyA logical topology is how devices appear connected to theuser. A physical topology is how they are actually intercon-nected with wires and cables.


LOMMFLaser-optimized 50/125 µm multimode fiber where thebandwidth is optimized at 850 nm wavelength in supportof ≥ 1 Gb/s operation.

Loose Tube CableType of cable design whereby colored fibers are encasedin buffer tubes.

LossReduction in optical power due to adsorption, scatteringand/or reflection.

MACsAcronym for moves, adds and changes. Usually associatedwith data centers.

Main Cross-Connect (MC)The centralized portion of the backbone cabling usedto mechanically terminate and administer the backbonecabling, providing connectivity between equipment rooms,entrance facilities, horizontal cross-connects and interme-diate cross-connects.

Main Distribution Area (MDA)The space in a computer room where the maincross-connect is located.

Mass SplicingJoining two to 12 fibers simultaneously by fusing thefibers together.

Material DispersionPulse dispersion caused by the variation in the speedof light with wavelength.

MDPEMedium density polyethylene; a type of plastic material,used as outside plant, commonly cable jackets.

Mechanical SplicingJoining two fibers together by permanent or temporarymechanical means (vs. fusion splicing or connectors) toenable a continuous signal. The CamSplice™ MechanicalSplice is a good example.

Media (Telecommunications)Wire, cable, or conductors used for telecommunications.

Megahertz (MHz)A unit of frequency that is equal to one million cyclesper second.

Mesh NetworkA communications network in which there are at least twopathways to each node. A “fully meshed” network meansthat every node has a direct connection to every othernode, which is a very elaborate and expensive architecture.Most mesh networks are partially meshed and requiretraversing nodes to go from each one to every other.

MeterDevice to measure optical power level (dBm).

Meter/Foot MarksThe distance markings stamped on the cable jacket by thefactory in either m or ft.

Micrometer (µm)One millionth of a meter; 10-6 m; typically used to expressthe geometric dimension of fibers, e.g., 62.5 µm.

ModeA term used to describe an independent stable light pathin a fiber, as in multimode or single-mode.

Mode ConditionerThe practice of wrapping a multimode fiber around amandrel for the purpose of causing light in the cladding(cladding modes) to be lost, as well as to facilitate a moreeven distribution of light across the core.

Mode Field Diameter (MFD)The area of a single-mode fiber in which light actuallytravels. This is typically larger than the core of the fiber.

ModulationCoding of information onto the carrier frequency.This includes amplitude, frequency or phase modulationtechniques.

Motion Pictures Experts Group (MPEG)An ISO/ITU standard for compressing video.

MTP® ConnectorMTP is a registered trademark of USConec, Ltd. andrefers to an enhanced MPO-style connector containinga linear array of 12F.

Multi-fiber CableAn optical fiber cable that contains two or more fibers.

Multimode FiberAn optical waveguide in which light travels in multiplemodes. Typical core/cladding sizes (measured inmicrometers) are 62.5/125 and 50/125.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

MultiplexCombining two or more signals that can be individuallyrecovered into a single bit stream.

MultipointRefers to a communications line (network) that provides apath from one location to many.

Multi-StageAn architecture in which I/O and cross-connect switchesare used to increase fabric bandwidth, throughput andresilience.

Multi-TrunkingIncreasing available bandwidth by connecting more thanone switched port to a single device.

Multiuser Telecommunications OutletA telecommunications outlet used to serve more thanone work area, typically used in open-systems furnitureapplications.

National Electrical Code® (NEC®)Provides practical safeguarding of persons and propertyfrom hazards arising from the use of electricity. This codeis updated by the NEC® every three years.

NICNetwork interface card

OFLOver filled launch, typical of LED source systems.

On-the-Reel TestTest of a new reel of cable prior to installation to verifylength and condition of the fiber.

Optical FiberSee Fiber.

Optical HardwareHousings designed to facilitate splicing and/or terminationof optical fiber cable.

Optical Time Domain Reflectometer (OTDR)An instrument that measures the transmission characteristicsof optical fiber by sending a series of short pulses of lightdown the fiber and providing a graphic representation ofthe backscattered light.

Optical SkewThe difference in propagation time between multi-fibersof a parallel transmission system.

Optical WaveguideSee Fiber.

Open Systems Interconnection (OSI)Refers to a seven-layered model that serves as a guidelinefor creating and implementing network standards, devicesand Internet working schemes to allow communicationbetween multiple network devices.

OSEOptical splice enclosure

Parallel Optic TransmissionThe simultaneous transmission of related signal elementsover two or more separate fibers. Parallel optics relies onspatial division multiplexing, in which a signal is spatiallydivided among multiple fibers and simultaneouslytransmitted across those fibers.

Patch PanelA collection of connector panels located in a commonhousing.

PBXPrivate branch exchange. A private telecommunicationsswitching system.

PCHPretium® Connector housing

Physical MeshEach switch is connected directly to each of the otherswitches on the network.

Physical RingA cable layout in which each node is connected to two adja-cent nodes. There is not a central point of cable termination.

Physical StarA cable layout in which all cables route back to a centrallocation, directly or through other consolidation points.

PigtailOptical fiber cable that has connectors installed on oneend. See Cable Assembly.


PIN DiodeA semiconductor device used to convert optical signalsto electrical signals in a receiver.

PlenumAn air-handling space such as that found above drop-ceilingtiles or in raised floors; also, a fire code rating for indoorcable suitable for use in plenum spaces.

Plug & Play™ Universal SystemsA fiber optic preterminated cabling system designed forthe private networks environment. This innovative systemreduces installation time and cost, for both end users andcontractors, by offering factory-terminated cables andpolarity management. The modular design guaranteescompatibility, flexibility and system performance for alloptical connection spans.

PMDPhysical media dependent

Point-to-Point (P2P)Refers to a communications line that provides a path fromone location to another (point A to point B).

PolarityFiber positioning convention that maintains the transmitand receive signals over the entire link.

Polyethylene (PE)A type of plastic material used for outside plant cable jackets.

Polyvinyl Chloride (PVC)A common plastic used for insulating and jacketing manyinside and indoor/outdoor cable products; typically used inflame-retardant cables.

Polyvinylidene Fluoride (PVDF)A type of material used for cable jacketing, typically usedin plenum-rated cables.

PoETechnology that supplies power and communication to aremote device over the same cable, thus eliminating theneed for power cords.

PortThe transmit/receive connection that is found within a node.

Preconnectorized AssemblyA fiber optic cable that has been terminated by themanufacturer. The terminations can be housed in aprotective pulling grip allowing inner duct installation.The terminations can also be pre-installed in hardware.

PrefusingA low-current electric arc used to clean the fiber end priorto fusion splicing.

Pretium EDGE™A high-density preterminated optical cabling solutionthat simplifies installation and improves performancein the data center environment.

Pretium® SolutionsA subset of Corning Cable Systems LANscape® Solutions,the Pretium product solutions offers enhanced perform-ance or handling characteristics.

PSTNPublic switched telephone network

PulseWidthThe time duration of a laser pulse emitted by the OTDR;ranges from a few nanoseconds to 20 microseconds,depending on model. Short pulses provide higher resolu-tion for short cables, whereas longer pulses provide powerneeded to test long distance cables.

Quality of Service (QoS)Describes a network’s ability to send time-dependent data.

RackVertical support for equipment typically with 1.75-in ofspace between mounting holes. Standard rack sizes are19-in and 23-in wide.

Rack SpaceA unit of measure of 1.75-in for equipment space in a rack.Many housings are measured in rack space.

ReceiverAn electronic package that converts optical signals toelectrical signals.

ReferenceThe power level of the source as measured through atest jumper that will be connected to a fiber for testing.Measurements through the system fiber are compared tothis value and the difference is the system loss.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

ReflectanceThe ratio of reflected power to incident power at aconnector junction or other component or device, usuallymeasured in decibels and typically stated as a negativevalue, e.g., -30 dB. The terms return loss, back reflectionand reflectivity are also used synonymously to describedevice reflections, but are stated as positive values.

ReflectionLight which is reflected whenever there is a differencein media and the index of refraction, such as a connectorinterface, where air (different index) is present, or the endof a fiber, where glass meets air.

RepeaterA device used to regenerate an optical signal to allow anincrease in the system length.

ResilienceA network's ability to preserve in the presence of failures.Example: The mesh architecture offers multiple pathsbetween switches, so if a switch fails, only the users onthat switch are out of operation; all other users are stillfunctional.

Restricted Mode Launch (RML) BandwidthA test method for measuring the laser bandwidth of multi-mode fibers; detailed in TIA/EIA-455-204 (FOTP-204).Method is used to simulate launch characteristics of 1 GbEsystems.

Return LossSee Reflectance.

RFIRadio frequency interference

RHRelative humidity

RIORuggedized information outlet

RiserPathway for indoor cables that passes between floors,normally a vertical shaft or space; also, a fire-code ratingfor indoor cable suitable for use in riser spaces.

Route DiagramA schematic diagram showing the physical location/layoutof the fiber run and the location of splices and terminationpoints.

RouterProvides connection over the OSI network layer (layer 3)based on the IP address.

Round-Trip Delay (RTD)Twice the time required for a packet to travel acrossa network.

ScatteringThe loss of signal power (light) from the fiber corecaused by impurities or changes in the index of refractionof the fiber.

SCFSplice closure family

SequenceOne or more subsets of an exchange.

Serial Optic TransmissionThe sequential transmission of signal elements of a datagroup. The characters are transmitted in a sequence overa single fiber, rather than simultaneously over two or morefibers, as in parallel transmission.

Signature TraceAn OTDR trace that is scaled so the entire fiber run isvisible on the graph; traces meant to document a fiber aretypically set up and saved/printed in this fashion.

Simplex ConnectorSingle connector terminating a single strand of fiber.

Single-Mode Fiber (SMF)An optical waveguide (or fiber) in which the signal travelsin only one mode. The fiber has a small mode fielddiameter, typically around 9 m.

SNMPSimple network management protocol

SONETSynchronous optical network

SourceStabilized light-emitting device (LED or Laser) used witha meter to measure attenuation.


Space Division MultiplexingA method used to increase the data rate capacity betweentwo points by transmitting data over multiple differentchannels simultaneously. A single input signal is brokeninto many segments, each having very short duration.Each segment is transmitted over a separate physical chan-nel to the receive end. At the receive end, the segments arecombined back in the correct order into a single data string.

SPHSingle-panel housing

Splice ClosureA container used to house cable splice points and organizeand protect splice trays; typically used in outside plantenvironments.

Splice TraysSplice trays are required in order to protect, store andorganize fibers and splices at splice points. A splice tray istypically a thin, rectangular sheet metal or plastic tray basewith a splice organizer, which has a removable sheet metalor plastic cover.

SplicingJoining of bare fiber ends to one another. See FusionSplice and Mechanical Splicing.

Star ToplogyA topology in which telecommunications cables aredistributed from a central point.

Step IndexA fiber that has a constant index of refraction for thecladding as well as the core. It is called step index becausethe index of refraction profile resembles a step.

Storage Area Network (SAN)A high-speed network that uses the Fibre Channeltransmission protocol to interconnect different kindsof data storage devices with associated data servers onbehalf of a larger network of users.

STPShielded twisted-pair

Super Absorbent Polymer (SAP)Hydrophilic polyacrylates (water absorbing plastics, babydiaper technology) that are used in state-of-the-art water-blocked cables. These plastics are adhered to tapes or yarnsin a cable to replace 100-year-old grease waterblockingtechnology.

Telecommunications Room (TR)An enclosed space for housing telecommunicationsequipment, cable terminations and cross-connects.The TR is the recognized cross-connect between thebackbone and horizontal cabling.

TerminationA method of preparing a fiber end for quick connectionto another fiber or device; involves use of a fiber opticconnector.

Test JumperA short, 2-3 m jumper used with meter/source for bothreferencing as well as conveniently connecting to eachconnector in a patch panel.

Through SpliceA splice used to join similar cables. This can be done toextend the length of a cable or distribute fiber circuits tosmaller count cables.

TIATelecommunications Industry Association

Tight-Buffered CableType of cable construction in which each glass fiber istightly buffered by a protective thermoplastic coatingto a diameter of 900 µm, providing ease of handling andconnectorization.

TopologyThe physical or logical arrangement of a telecommunicationssystem i.e. Star, Ring or Mesh.

TraceThe OTDR’s graphical representation of a fiber whichdisplays relative power on the vertical and distance onthe horizontal scales.

Transition SpliceA splice, usually in the building entrance, to joinflame-rated and non-flame-rated cables together.


INTRO

DUCTIO

NTO

DATACEN

TERSDESIG

NINGTHEPHYSICAL

INFRASTRU

CTURE

DEPLO

YINGTHEPHYSICAL

INFRASTRU

CTURE

PERFORM

ANCE

METRICS

ANDADMINISTRATIO

NINFORM

ATIONAN

DTOOLS

Transmission Control Protocol/InternetProtocol (TCP/IP)Four-layer communication protocol developed by theU.S. Government.

TransmitterAn electronic device used to convert an electrical informa-tion signal to a corresponding optical signal for transmissionby fiber. Transmitters are typically light emitting diodes(LEDs), VCSELs or laser diodes.

UDPUser datagram protocol

Ultraviolet Cure ConnectorA field-installable connector with a polymer epoxy thathardens when exposed to ultraviolet light.

UniCam® ConnectorCorning Cable Systems’ field-installable connector thatrequires no epoxy and no polishing.

Uplink PortA port on a network hub or switch that is used to connectto other hubs and switches rather than an end station.

UTPUnshielded twisted-pair

Vertical Cavity Surface Emitting Laser (VCSEL)Vertical cavity surface emitting laser. Pronounced “vixel”,VCSEL is a type of laser diode that emits light from itssurface rather than its edge. A VCSEL’s circular beam iseasy to couple with a fiber and due to its surface-emissionarchitecture, can be tested on the wafer. VCSELs are alsonoted for their excellent power efficiency and durability.

Visual Fault Locator (VFL)A visible Class II red light laser, typically 630-670 nm,which is used to check short cables such as pigtails andjumpers for breaks by causing the break to glow red.

VoIPVoice over Internet protocol

Waveguide DispersionDispersion caused by light traveling in the cladding of thesingle-mode fiber.

WavelengthThe distance between two successive points of an electro-magnetic waveform, usually measured in nanometers (nm).

WCHWall-mountable connector housing

WCH-SSHWCH slack storage housing (mounts behind the WCH)

WDMWavelength division multiplexing. The simultaneoustransmission of two or more wavelengths of light ona single fiber.

WICWall-mountable interconnect center

WMOWorkstation multimedia outlet

Work Area Telecommunications OutletA connecting device located in a work area at which thehorizontal cabling terminates and provides connectivityfor work area patch cords.

WSHWall-mountable splice housing

Zone Distribution Area (ZDA)A space in a computer room where a zone outlet or aconsolidation point is located.

Zero DispersionWavelengthWavelength at which the chromatic dispersion of anoptical fiber is zero.


Corning Cable Systems LLC • PO Box 489 • Hickory, NC 28603-0489 USA800-743-2675 • FAX: 828-325-5060 • International: +1-828-901-5000 • www.corning.com/cablesystemsCorning Cable Systems reserves the right to improve, enhance and modify the features and specifications of Corning Cable Systems products without prior notification. ALTOS, LANscape,Pretium and UniCam are registered trademarks of Corning Cable Systems Brands, Inc. CamSplice, LID-SYSTEM, Plug & Play and Pretium EDGE are trademarks of Corning CableSystems Brands, Inc. ClearCurve and Corning are registered trademarks of Corning Incorporated. MTP is a registered trademark of USConec, Ltd. All other trademarks are the propertiesof their respective owners. Corning Cable Systems is ISO 9001 certified. © 2010 Corning Cable Systems. All rights reserved. Published in the USA. LAN-1160-EN / November 2010

Guía de Diseño Del Centro de Datos

Documents

Transcript of Guía de Diseño Del Centro de Datos