【正文】 這嚴(yán)重限制了總帶寬。 端口緩沖內(nèi)存 通過采用端口緩沖內(nèi)存，交換機 (例如 Catalyst 5500)能夠為每個以太網(wǎng)端口提供一定數(shù)量的高速內(nèi)存，這些內(nèi)存可用于幀發(fā)送之前的幀緩沖。對于采用 Supervisor 1的 Catalyst 4000系列，它采用 8MB 的 SRAM（ Static RandomAccess Memory，靜態(tài)隨機訪問內(nèi)存）作為動態(tài)幀緩沖區(qū)。在早期的共享總線環(huán)境中，中央仲裁器采用循環(huán)服務(wù)的方法在不同的線路卡之間移動流量。在正常工作的情況下，因為交換總線能夠以非常高的速度為幀提供服務(wù)，所以只需要小的輸入隊列。 6. 數(shù)據(jù)轉(zhuǎn)發(fā) 無論交換矩陣類型如何，交換機都必須決定哪些端口轉(zhuǎn)發(fā)幀和哪些端口應(yīng)當(dāng)清空或者丟棄幀。為了獲得交換總線的訪問，每個線路卡的本地仲裁器都必須等待中央仲裁器的分配次序。 Catalyst 4500 Supervisor IV 中英文資料 11 采用 16MB 的 SRAM 用于分組緩沖區(qū)。為了最大限度利用緩沖的優(yōu)勢，方法之一是采用靈活的緩沖區(qū)尺寸。對于不同的交換機平臺，術(shù)語交叉矩陣意味著不同的內(nèi)容，但基本都指線路卡之間能夠同時使用多個數(shù)據(jù)信道或者通路。 圖 22 循環(huán)服務(wù)順序 圖 23 說明了共享總線體系結(jié)構(gòu)中將接收端口或入口處的幀移動到發(fā)送端口或出口的基本原理 ,其中各步驟說明如下。為什么交換機檢查幀的前64 個字節(jié)呢 ?因為在設(shè)計良好的以太網(wǎng)網(wǎng)絡(luò)中 ,碰撞碎片必須在前 64 字節(jié)中檢測出來。隨后，本章介紹保證高效數(shù)據(jù)交換的一些機制。 1. 數(shù)據(jù)接收 交換模式 在局域網(wǎng)交換中，根據(jù)交換機功能的不同，第一步就是從發(fā)送設(shè)備或主機接收幀或分組。 中英文資料 8 圖 21 比較各種交換模式之間的區(qū)別。如果幀存在缺陷（例如殘幀、碎片、無效 CRC 或者巨型幀），那么端口將丟棄該幀，并且將增加相關(guān)計數(shù)器的數(shù)值。在 Cisco Catalyst 5500 平臺中，吉比特以太網(wǎng)線路卡要求訪問所有 3 條數(shù)據(jù)總線。 通過使用 168KB 的發(fā)送緩沖區(qū)，每個端口最多能夠創(chuàng)建 2500 個 64 字節(jié)的緩沖 區(qū)。 4. 過度預(yù)定交換矩陣 交換機制造商喜 歡使用術(shù)語無阻塞米表明連接到交換矩陣的交換端口能夠達(dá)到線速。 如圖 27 所示，流量發(fā)生器能夠產(chǎn)生 擁塞現(xiàn)象。每臺 Catalyst 交換機都會為每個分組創(chuàng)建報頭或者標(biāo)簽，并且以該報頭或者標(biāo)簽作為轉(zhuǎn)發(fā)決策的基礎(chǔ)。 Catalyst 交換機采用多種技術(shù)來避免線端阻塞，例如采用每端口緩沖。為了確定交換矩陣的總連接帶寬，請查看各種平臺和線路卡的技術(shù)指標(biāo)。 Catalyst 1200 能夠支持以太網(wǎng)和 FDDI，并且具有 4MB 的共享分組 DRAM（ Dynamic RandomAccess Memory，動態(tài)隨機訪問內(nèi)存）。 如果發(fā)往端口的流量超過了它所能發(fā)送的流量 ,那么就需要使用緩沖。 2在這個示例中，假定交換機知道主機 2 的位置，并且僅在連接到主機 2 的端口發(fā)送幀。根據(jù)交換機配置的情況，仲裁器能夠使用多種公平方法。 直通模式 如果交換機工作在直通模式，那么它將只接收和檢查幀的的前 6 個字節(jié)。s examination of switching begins from a Layer 2 point of view. Depending on the model, varying amounts of each frame are stored and examined before being switched. Three types of switching modes have been supported on Catalyst switches: ?Cut through ?Fragment free ?Store and forward These three switching modes differ in how much of the frame is received and examined by the switch before a forwarding decision is made. The next sections describe each mode in detail. CutThrough Mode Switches operating in cutthrough mode receive and examine only the first 6 bytes of a frame. These first 6 bytes represent the destination MAC address of the frame, which is sufficient information to make a forwarding decision. Although cutthrough switching offers the least latency when transmitting frames, it is susceptible to transmitting fragments created via Ether collisions, runts (frames less than 64 bytes), or damaged frames. FragmentFree Mode Switches operating in fragmentfree mode receive and examine the first 64 bytes of frame. Fragment free is referred to as fast forward mode in some Cisco Catalyst documentation. Why examine 64 bytes? In a properly designed Ether work, collision fragments must be detected in the first 64 bytes. StoreandForward Mode Switches operating in storeandforward mode receive and examine the entire frame, resulting in the most errorfree type of switching. As switches utilizing faster processor and applicationspecific integrated circuits (ASICs) were introduced, the need to support cutthrough and fragmentfree switching was no longer necessary. As a result, all new 中英文資料 2 Cisco Catalyst switches utilize storeandforward switching. Figure21 pares each of the switching modes. Modes 2. Switching Data Regardless of how many bytes of each frame are examined by the switch, the frame must eventually be switched from the input or ingress port to one or more output or egress ports. A switch fabric is a general term for the munication channels used by the switch to transport frames, carry forwarding decision information, and relay management information throughout the switch. A parison could be made between the switching fabric in a Catalyst switch and a transmission on an automobile. In an automobile, the transmission is responsible for relaying power from the engine to the wheels of the car. In a Catalyst switch, the switch fabric is responsible for relaying frames from an input or ingress port to one or more output or egress ports. Regardless of model, whenever a new switching platform is introduced, the documentation will generally refer to the transmission as the switching fabric. Although a variety of techniques have been used to implement switching fabrics on Cisco Catalyst platforms, two major architectures of switch fabrics are mon: ?Shared bus ?Crossbar Shared Bus Switching In a shared bus architecture, all line modules in the switch share one data path. A central arbiter determines how and when to grant requests for access to the bus from each line card. Various methods of achieving fairness can be used by the arbiter depending on the configuration of the switch. A shared bus architecture is much like multiple lines at an airport ticket counter, with only one ticketing agent processing customers at any given time. Figure22illustrates a roundrobin servicing of frames as they enter a switch. Roundrobin is the simp