Introduction

The accelerating growth of Local Area Network (LAN) traffic is pushing network administrators to look to higher-speed network technologies to solve the bandwidth crunch. Highly reliable networks are critical to the success of the enterprise, so ease of installation and support are primary considerations in the choice of network technology. Since the introduction in 1986 of star-wired 10BASE-T hubs, structured wiring systems have continued to evolve and hubs and switches have become increasingly reliable. Today, Ethernet networks are rapidly approaching the reliability level associated with their telephone ancestors, and are relatively simple to understand..

The Fast Ethernet standard, approved in 1995, established Ethernet as a scalable technology. Now, the development of Gigabit Ethernet extends the scalability of Ethernet even further. Independent market research has indicated a strong interest among network users in adopting Gigabit Ethernet technology, specifically Fast Ethernet hubs and switches with Gigabit Ethernet uplinks, Gigabit Ethernet switches and repeaters, and Gigabit Ethernet server NIC. The goal of the IEEE 802.3z Task Force, which developed the Gigabit Ethernet standard, was to specify connections that delivered 10 times the performance of Fast Ethernet at very affordable prices. Because Gigabit Ethernet leverages existing Ethernet technologies, it also leverages Ethernet’s fiercely competitive industry cost curve. Gigabit Ethernet products are a good value today. Currently, switched 1000 Mbps Gigabit Ethernet provides the best price performance of all the leading high-speed LAN technologies.

The initial applications for Gigabit Ethernet will be for campuses or buildings requiring greater bandwidth between routers, switches, hubs, repeaters and servers. In its early phase, Gigabit Ethernet is not expected to be deployed widely to the desktop. In all scenarios the network operating system (NOS), applications and NIC drivers (Network Interface Card) at the desktop will remain unchanged. The MIS manager can also leverage not only the existing multimode fiber, but also the current investment in network management applications

Ethernet Milestones

One of the key issues for gigabit Ethernet is the maximum size of the network. The best way to understand the limits and the reasons for them is to take a quick look at some Ethernet history.

1980 : The original IEEE 802.3 Ethernet defined a mechanism called Carrier-Sense Multiple Access with Collision Detection (CSMA/CD). This scheme ensures that all stations are granted access on a first-come, first-served basis. Since Ethernet was intended only to carry data, no provisions were made for quality of service or prioritization. CSMA/CD simply ensures that the same access rules apply equally to all network nodes. The designers of Ethernet worked it out so that stations up to 2 kilometers apart could sense when a collision occurs. This distance limitation results from the relationship between the time required to transmit a minimum-sized Ethernet frame 64 bytes and the ability to detect a collision. When a collision occurs, the MAC layer detects it and sends a jam signal, telling the transmitting stations to back off and then retry.

1994: IEEE defined IEEE802.3u for 100Base-T . It maintained the Ethernet framing format and raised the speed limit to 100 Mbit/s. But increasing the clock rate tenfold means that the time needed to transmit a frame is reduced by a factor of ten. That, in turn, directly affects network diameter, shrinking it from 2 kilometers for 10Base-T to 200 meters for 100Base-T.

1995 : Fast Ethernet standard established Ethernet as a scalable technology

1996 : More than 83 percent of all installed network connections were according to IDC. This represents over 120 million interconnected PCs, workstations and servers. The remaining network connections are a combination of Token Ring, Fiber Distributed Data Interface (FDDI), Asynchronous Transfer Mode (ATM) and other protocols. All popular operating systems and applications are Ethernet-compatible, as are upper-layer protocol stacks such as Transmission Control Protocol/Internet Protocol (TCP/IP), IPX, NetBEUI and DECnet.

1998 : Gigabit Ethernet standard established Ethernet as a fasterst technology.

1999 : Milestone year for Ethernet network equipment as the technology captures 77 percent of shipments. IDC projects Ethernet network interface card (NIC) shipments will exceed 42 million units, and Ethernet hub shipments will exceed 54 million ports. In contrast, ATM, FDDI/CDDI and Token Ring network interface card shipments combined are expected to reach just 5 million in 1997, 12 percent of the total. ATM, FDDI/CDDI and Token Ring hub ports are expected to be 7 million, 13 percent of the total. IDC projects that Ethernet dominance will continue beyond 1999. Several factors have contributed to making Ethernet the most popular network technology in use today.

Gigabit Ethernet: Meeting Tomorrow's Needs

Like its Ethernet and Fast Ethernet precursors, Gigabit Ethernet is a physical (PHY) and media access control (MAC) layer technology, specifying the Layer 2 data link layer of the OSI protocol model. It complements upper-layer protocols TCP and IP, which specify the Layer 4 transport and Layer 3 network portions and enable reliable communication services between applications.

Gigabit Ethernet is being developed to support switch-to-switch and switch-to-server connections in building and campus backbones, the standards effort has devoted attention to developing specifications for fiber optic transceivers, optics and distance goals over multimode and single-mode fiber. For multimode fiber, the goal is to reach distances of 260 meters (850 ft.) on 62.5-µm-diameter fiber and 550 meters (1,805 ft.) on 50-µm-diameter multimode fiber.

Transceivers and Media for single-mode fiber, the goal is to reach at least 3 km (1.8 mi.)

The network topology for Gigabit Ethernet follows the traditional rules of Ethernet. At Layer 2, the Spanning Tree Protocol is used to ensure that there are no logical loops in the network, creating a hierarchical tree topology. More complex LAN configurations, including parallel data paths, are created through the use of routing technology. At any speed, Ethernet is designed for transporting data in the local area environment. Like Ethernet and Fast Ethernet, Gigabit Ethernet leverages other technologies and standards to provide higher-level services such as Class of Service (CoS) traffic prioritization and Quality of Service (QoS) data delivery that limit jitter and latency.

The Gigabit Ethernet Standard

The 1000BASE-X (IEEE 802.3z) Gigabit Ethernet standard was ratified in June 1998, after more than two years of intense effort within the IEEE 802.3 Ethernet committee. The key objective of the 802.3z Gigabit Ethernet Task Force was to develop a Gigabit Ethernet standard that encompassed the following:

1. Allowed half- and full-duplex operation at speeds of 1000 Mbps
2. Used the 802.3 Ethernet frame formats
3. Used the CSMA/CD access method with support for one repeater per collision domain
4. Addressed backward compatibility with 10BASE-T and 100BASE-T technologies

Because the fundamental features of the 802.3z specification have been stable during the last stages of the standardization process, network vendors have been able to build and deliver quality, mature products to the marketplace for many months. In addition, numerous interoperability demonstrations have been sponsored by the Gigabit Ethernet Alliance and other independent organizations, giving customers confidence in using Gigabit Ethernet products in their production networks.

Gigabit Ethernet : Key criteria for High-Speed Network

Although each network faces different issues, Gigabit Ethernet meets several key criteria for choosing a high-speed network:

Easy Migration to Higher Performance

One of the most important questions network administrators face is how to get higher bandwidth without disrupting the existing network. Gigabit Ethernet follows the same form, fit and function as its 10 Mbps and 100 Mbps Ethernet precursors, allowing a straightforward, incremental migration to higher-speed networking. All three Ethernet speeds use the same IEEE 802.3 frame format, full-duplex operation and flow control methods. In half-duplex mode, Gigabit Ethernet employs the same fundamental CSMA/CD access method to resolve contention for the shared media. And, Gigabit Ethernet uses the same management objects defined by the IEEE 802.3 group. Gigabit Ethernet is Ethernet, only faster.

Ethernet Frame Format 

In terms of data rate, Gigabit Ethernet simply moves the decimal point, increasing the Fast Ethernet transmission rate tenfold, from 100 Mbps to 1000 Mbps Gigabit Ethernet is Ethernet that provides speeds of 1000 Mbps. It uses the same Ethernet frame format and media access control technology as all other 802.3 Ethernet technologies. It also uses the same 802.3 full-duplex Ethernet technology and 802.3 flow control.

It is simple to connect existing lower-speed Ethernet devices to Gigabit Ethernet devices using LAN switches or routers to adapt one physical line speed to the other. Gigabit Ethernet uses the same variable-length (64- to 1514-byte packets) IEEE 802.3 frame format found in Ethernet and Fast Ethernet (Figure 1). Because the frame format and size are the same for all Ethernet technologies, no other network changes are necessary. This evolutionary upgrade path allows Gigabit Ethernet to be seamlessly integrated into existing Ethernet and Fast Ethernet networks. In contrast, other high speed technologies use fundamentally different frame formats. High-speed ATM, for example, implements a fixed-length data cell. When connecting Ethernet and Fast Ethernet to ATM, the switch or router must translate each ATM cell to an Ethernet frame, and vice versa.

Full and Half-Duplex Operation    

As defined by the IEEE 802.3x specification, two nodes connected via a full-duplex, switched path can simultaneously send and receive packets. Gigabit Ethernet follows this standard to communicate in full-duplex mode. Gigabit Ethernet also employs standard Ethernet flow control methods to avoid congestion and overloading. When operating in half-duplex mode, Gigabit Ethernet adopts the same fundamental CSMA/CD access method to resolve contention for the shared media. The Gigabit Ethernet CSMA/CD method has been enhanced in order to maintain a 200-meter collision diameter at gigabit speeds. Without this enhancement, minimum-sized Ethernet packets could complete transmission before the transmitting station senses a collision, thereby violating the CSMA/CD method. To resolve this issue, both the minimum CSMA/CD carrier time and the Ethernet slot time have been extended from their present value of 64 bytes to a new value of 512 bytes. (Note that the minimum packet length of 64 bytes has not been affected.) Packets smaller than 512 bytes have an extra carrier extension. Packets longer than 512 bytes are not extended. These changes, which can impact small-packet performance, have been offset by incorporating a new feature, called packet bursting, into the CSMA/CD algorithm. Packet bursting will allow servers, switches and other devices to send bursts of small packets in order to fully utilize available bandwidth. Devices that operate in full-duplex mode (switches and buffered distributors) are not subject to the carrier extension, slot time extension or packet bursting changes. Full-duplex devices will continue to use the regular Ethernet 96-bit interframe gap (IFG) and 64-byte minimum packet size.

Management Objects    

As in the transition from Ethernet to Fast Ethernet, the fundamental management objects familiar to most network managers will be carried forward with Gigabit Ethernet. For example, SNMP defines a standard method to collect device-level Ethernet information. SNMP uses management information base (MIB) structures to record key statistics such as collision count, packets transmitted or received, error rates and other device-level information. Additional information is collected by remote monitoring (RMON) agents to aggregate the statistics for presentation via a network management application. Because Gigabit Ethernet uses standard Ethernet frames, the same MIBs and RMON agents can be utilized to provide network management at gigabit speeds.

Low Cost of Ownership

Cost of ownership is an important factor in evaluating any new networking technology. The overall cost of ownership includes not only the purchase price of equipment, but also the cost of training, maintenance and troubleshooting. Competition and economies of scale have driven the purchase price of Ethernet connections down significantly. Though Fast Ethernet products have been shipping only since 1994, even these products have experienced significant price declines over the past two years. Gigabit Ethernet will follow the same price trends as Fast Ethernet. Even early products will provide cost-effective connections for gigabit transmission rates. The IEEE’s goal is to provide Gigabit Ethernet connections at two to three times the cost of a 100BASE-FX interface. As volume builds, reduced line width IC processes are implemented and low-cost opto-electronic devices are developed, the cost of Gigabit Ethernet interfaces will decline. Switched Gigabit Ethernet connections are expected to be lower in cost than 622 Mbps ATM interfaces (assuming identical physical media interfaces), because of the relative simplicity of Ethernet and higher shipment volumes. Gigabit Ethernet repeater interfaces will be significantly lower in cost than 622 Mbps ATM connections, providing users with cost-effective alternatives for data center network backbone and server connections. Table 3 illustrates current prices for Ethernet, Fast Ethernet, FDDI and 622 Mbps ATM multimode and the target range for Gigabit Ethernet based on the IEEE 802.3z goals. Over time, advances in silicon, including 0.35-micron CMOS ASIC technology, will provide even greater performance gains and cost reduction opportunities that will result in a new, even more cost-effective generation of Ethernet technology. Current 0.5-micron technology can accommodate about 0.5 million gates and is limited to transmission rates of about 500 Mbps. Analysis indicates that 0.35-micron processes will achieve 1250 Mbps operation and economically fit one million gates on a single die. This is more than enough to fit a complete Ethernet switch, including management, a significant amount of buffer memory, and an embedded 32-bit controller, on a single die—with obvious cost advantages. Finally, because the installed base of users is already familiar with Ethernet technology, maintenance and troubleshooting tools, the support costs associated with Gigabit Ethernet will be far lower than other technologies. Gigabit Ethernet requires only incremental training of personnel and incremental purchase of maintenance and troubleshooting tools. In addition, deployment of Gigabit Ethernet will be faster than alternative technologies. Once upgraded with training and tools, network support staff will be able to confidently install, troubleshoot and support Gigabit Ethernet installations

Support for New Applications and Data Types

1. Increased bandwidth provided by Fast Ethernet and Gigabit Ethernet, enhanced by LAN switching
2. The emergence of new protocols, such as Resource Reservation Protocol (RSVP), that provide bandwidth reservation
3. The emergence of new standards such as 802.1Q and/or 802.1p which will provide virtual LAN (VLAN) and explicit priority information for packets in the network
4. The widespread use of advanced video compression such as MPEG-2

Flexible Internetworking and Network Design

Gigabit Ethernet Cabling and Distance Specifications

The physical (PHY) layer is a crucial part of the Gigabit Ethernet specification. It provides the interface between the media access control (MAC) layer and the transceivers in Gigabit Ethernet hardware. The PHY layer performs encoding, decoding, carrier sense, and link monitor functions.

The IEEE 802.3z Gigabit Ethernet standard ratified in June 1998 includes three physical layer specifications :

1. Two for fiber optic media—1000BASE-SX and 1000BASE-LX.
2. One for shielded copper media—1000BASE-CX.
3. Defining the physical layer to run Gigabit Ethernet over the installed base of Category 5 unshielded twisted pair (UTP) cabling. This 1000BASE-T effort is scheduled for completion in June 1999.

Specifications for fiber cabling

1. 1000BASE-SX
2. "S" for short wavelength) defines optical transceivers or physical layer devices for laser fiber cabling. The wavelength of SX lasers is commonly referred to as 850 nm. 1000BASE-SX is based on the Fibre Channel signaling specification and uses the 8B/10Bphysical coding sublayer (PCS). Targeted for multimode fiber only, 1000BASE-SX transceivers are less costly than those found in products implementing the long wavelength specification.
3. 1000BASE-LX
4. "L" for long wavelength defines optical transceivers or physical layer devices for laser fiber cabling. The wavelength of LX lasers is 1270 to 1355 nm (commonly referred to as 1350 nm). 1000BASE-LX is also based on the Fibre Channel signaling specification and uses FC 8B/10B physical coding sublayer (PCS). 1000BASE-LX is specified for use on either multimode or single-mode fiber.

   Transceiver	Fiber Diameter (microns)	Bandwidth (MHz*km)	Minimum Range (meters)
   1000BASE-SX	MM 62.5 MM 62.5 MM 50 MM 50	160 200 400 500	2–220 2–275 2–500 2–550
   1000BASE-LX	MM 62.5 MM 50 MM 50 SM 9	500 400 500 NA	2–550 2–550 2–550 2–5000

   The distance Gigabit Ethernet can reach depends on the bandwidth (measured in MHz*km)—the greater the bandwidth of the fiber, the further the distance supported. It’s also important to note that IEEE specifies minimum rather than maximum ranges, and under average operating conditions, the minimum specified distance can be exceeded by a factor of three or four. However, most network managers are conservative when they design networks and use the IEEE specifications as the maximum distances.

5. 1000BASE-LH

Copper cabling specifications

1. 1. 1000BASE-CX
   2. "C" for copper - defines transceivers or PHY layer devices for shielded copper cabling. 1000BASE-CX is intended for short-haul copper connections (25 meters or less) within wiring closets. In 1998, no vendor had announced plans to develop 1000BASE-CX products. This physical interface uses the Fibre Channel 8B/10B physical coding sublayer (PCS), a different driver on shielded copper, and 150 ohm balanced cable, and it supports wiring distances up to 25 meters. (Note that IBM Type 1 cable is specifically mentioned in the IEEE 802.3z standard as not recommended.)
   3. 1000BASE-T

Implementing Gigabit Ethernet in the Network

Assessing a new technology like Gigabit Ethernet involves many facets, including comparing it to other technologies and evaluating its impact on the existing network topology and network equipment. The topics discussed here are intended to help network managers consider both the advantages and limitations of

Gigabit Ethernet.

Link or port aggregation or trunking often comes up in discussions of Gigabit Ethernet migration. Link aggregation is the ability to support multiple, point-to-point, parallel active links between switches or between a switch and a server. The advantages of link aggregation are higher bandwidth, redundant links, and load sharing. Link aggregation and Gigabit Ethernet are complementary technologies. If the business requirement is simply to add more band-8 width between devices, and the devices can be upgraded to Gigabit Ethernet, Gigabit Ethernet is the right choice. If the business requirement includes the need for resilient and redundant links with load balancing, then link aggregation (aggregating multiple links into one logical connection) is the right choice.

Gigabit Ethernet links can of course themselves be aggregated with link aggregation technology. The primary application of link aggregation technology in the near future will be to build resilient, redundant links between Layer 2 and Layer 3 Gigabit Ethernet switches. Many vendor-specific link aggregation implementations exist in the industry today. For example, 3Com supports link aggregation for FDDI, Fast Ethernet, and Gigabit Ethernet. And in keeping with its standards-based solution strategy, 3Com helped initiate the IEEE 802.3ad standards effort for link aggregation in 1998.

Optimizing Servers to Handle Gigabit Ethernet

While Gigabit Ethernet relieves the bottleneck at the server, server environments are not yet fully optimized to handle the entire available gigabit bandwidth. The good news is that server systems and their software and hardware components are rapidly evolving to ensure that the server system bandwidth capacity will handle the Gigabit Ethernet data rate. In addition, Gigabit Ethernet network interface cards (NICs) currently under design will overcome some of the server system bottlenecks and be optimized for the upcoming evolution in server architecture. The PCI bus is the predominant bus in x86 platforms, and is also available in some non-x86 systems. Although the original PCI bus present in earlier server systems had insufficient bandwidth to carry gigabit-speed I/O traffic, in newer server systems the bus is quickly becoming wider and faster and is not a bottleneck for the Gigabit Ethernet network connection. The real bandwidth of the PCI bus is slightly lower than the maximum bus bandwidth due to the PCI bus overheads involved. At a minimum, the wider 64-bit bus at 33 MHz is required. In full-duplex mode, Gigabit Ethernet provides 2 Gbps bandwidth. Therefore, with the faster 66 MHz bus speed due out in the next few months, the PCI bus will no longer be the critical bottleneck in the system.

• At least one x86 non-Intel chipset server system supports 66 MHz PCI already.

Conclusion

Gigabit Ethernet is a natural extension of Ethernet networking. It brings not only increased bandwidth but also new features that enable it to function as an intelligent network, while keeping cost for support and implementation low. Better yet, the actual cost of Gigabit Ethernet products is projected to be lower than any other solution.

Ke Atas