Cisco 12008 Gigabit Switch Router Installation and Configuration Guide
Product Overview
Downloads: This chapterpdf (PDF - 1.09MB) The complete bookPDF (PDF - 5.95MB) | Feedback

Product Overview

Table Of Contents

Product Overview

Cisco's Next Generation of Routers

Features of the Cisco 12008 Router

Overview of the Cisco 12008

Router Enclosure

Cable-Management System

Card Cage Fan Tray

Power Supply Fan Tray

AC-Input and DC-Input Power Supplies

Operating Modes of the Power Supplies

Features of the Power Supplies

Characteristics of the Power Supplies

AC-Input Power Supply Faceplate

DC-Input Power Supply Faceplate

Upper Card Cage and Associated Components

Gigabit Route Processor

Performance Route Processor

Switch Fabric of the Cisco 12008

Clock and Scheduler Card

Cisco 12000 Series Line Cards

Air Filter Assembly

Lower Card Cage and Associated Components

Switch Fabric Cards

Power Distribution System in the Cisco 12008

Cisco 12008 Environmental Monitoring Facility

System Specifications

Agency Approvals


Product Overview


This chapter presents an overview of the Cisco 12008 Gigabit Switch Router.

The following sections are included in this chapter:

Cisco's Next Generation of Routers

Features of the Cisco 12008 Router

Overview of the Cisco 12008

Router Enclosure

Cable-Management System

Card Cage Fan Tray

Power Supply Fan Tray

AC-Input and DC-Input Power Supplies

Upper Card Cage and Associated Components

Air Filter Assembly

Lower Card Cage and Associated Components

Power Distribution System in the Cisco 12008

Cisco 12008 Environmental Monitoring Facility

System Specifications

Agency Approvals

Cisco's Next Generation of Routers

Cisco Systems' new family of Internet switching and routing products, referred to collectively as the Cisco 12000 Series Gigabit Switch Routers, consists of the following models:

Cisco 12016 Gigabit Switch Router—A 16-slot, carrier-class platform that supports Internet protocol (IP) switching capacity of up to 160 Gbps.

Cisco 12012 Gigabit Switch Router—A 12-slot version that supports IP datagram switching capacities ranging from 15 to 60 Gbps.

Cisco 12008 Gigabit Switch Router—An 8-slot version that supports IP datagram switching capacities ranging from 10 to 40 Gbps. The Cisco 12008 is the subject of this document.

The architecture of the Cisco 12000 Series Gigabit Switch Routers provides the following networking capabilities and features:

Scalable bandwidth—Supports high-speed transmission of IP datagrams through use of Cisco 12000 series line cards. The network interfaces reside on the line cards, providing connectivity between the router's switch fabric and external networks.

Scalable performance—Supports multi-gigabit bandwidth switching capacities ranging from 5 to 60 Gbps, providing high-performance support for IP-based networks and wide-area networks (WANs).

Scalable services—Supports sophisticated congestion management, multicast services, and quality-of-service (QOS) features.

Carrier-class design—Supports extensive SONET/Synchrononous Digital Hierarchy (SDH) integration; supports a hot-swapping capability for field-replaceable units (FRUs).

The networking capabilities and features of the Cisco 12000 series of routers make them ideally suited to meet the needs of the following classes of users:

Internet service providers (ISPs)

Carriers providing Internet services and utilities

Competitive access providers (CAPs)

Enterprise wide-area network (WAN) backbones

Metropolitan-area network (MAN) backbones

Features of the Cisco 12008 Router

The Cisco 12008 router incorporates a high-speed switching fabric that provides high data-handling capacities for IP-based local- and wide-area networks. is a front view of the Cisco 12008 router.

All of the router's major components and FRUs are accessible from the front of the router enclosure, making the router easy to install, configure, and maintain.

Figure 1-1 Cisco 12008 Gigabit Switch Router

The Cisco 12008 supports the following features:

Online insertion and removal (OIR) capability—This feature allows you to insert or remove the following router components:

Power supplies—One AC-input power supply or one DC-input power supply is a required router component. You can remove or replace a power supply, without disrupting system operations, only if a second (redundant) unit of the same type is installed in the system.

The power supplies of both types are hot-swappable, load-sharing units. In a system equipped with two AC-input power supplies or two DC-input power supplies, if one of the units fails or if the power source for one of the units fails, the surviving power supply continues to operate to sustain normal router operations.


Note   The Cisco 12008 does not support a mixture of AC-input and DC-input power supplies.


Cisco 12000 series line cards—Any line card supported by the Cisco 12008 router can be inserted into or removed from the router with no disruption to system operations.

However, the functions performed by the removed card are lost to the system temporarily until the card is either reinstalled or replaced by a like (and identically configured) line card.

Route Processor (RP)—As a required router component, an RP can be removed and replaced, but you must power down the router before doing so.

An RP must be installed and operational at all times for normal system operations to be sustained.

Clock and scheduler card (CSC)—Also a required component, a CSC can be removed and replaced, without disrupting normal system operations, only if a second (redundant) CSC is installed in the system.

One CSC must be present and operational at all times to maintain normal system operations.

Switch fabric card (SFC)—An optional set of three SFCs can be installed in the router at any time to provide additional switch fabric to the router. These cards increase the data handling capacity of the router.

Any one or all of the SFCs can be removed and replaced at any time without system operations being disrupted or the router being powered down.

For the length of time that any SFC is not functional, its switch fabric is lost to the router as a potential data path for the router's data handling and switching functions.

Separately orderable documents called configuration notes or replacement instructions are available for each of the FRUs described previously. These documents provide installation, removal, replacement, and configuration instructions for the FRUs.

Environmental monitoring system—The maintenance bus (MBus) facility of the Cisco  12008 functions as an environmental monitoring system for the router, enabling the router to monitor itself and alert site personnel to adverse electrical events or environmental conditions.

MBus software running in the RP, in combination with LEDs on the CSC faceplate, keep site personnel informed regarding the operational state of the router.

By signaling alarm conditions, such as component overheating or out-of-tolerance voltages, the router enables you to resolve adverse environmental conditions before operational limits are exceeded, thus preventing the router from shutting down.

The MBus facility of the router is described in greater detail in the section entitled "Cisco 12008 Environmental Monitoring Facility" on page 75.

Downloadable software—This feature allows you to remotely load new operational software into Flash memory on the RP without physically accessing the router. Thus, you can quickly, easily, and reliably perform software upgrades at any time.

Overview of the Cisco 12008

The Cisco 12008 is a modular system consisting of the elements shown in .

The following sections describe the major elements of the Cisco 12008 in greater detail.

Figure 1-2 Major Components of the Cisco 12008

Router Enclosure

The outer shell of the Cisco 12008 is a rigid, sheet metal structure with the following dimensions:

Width—17.4 inches (44.6 cm)

Depth—21.2 inches (54.4 cm)

Height— 24.8 inches (63.6 cm)

This enclosure, which houses all of the router's internal components, can be mounted in a telco rack or a four-post equipment rack, or the enclosure can be used as a freestanding unit.

The design of the enclosure permits front accessibility of all router components. All router components plug into a backplane that provides operating power for the components and interconnects them with each other.

The backplane, which is covered by a sheet metal panel that helps to completely enclose the rear of the router, incorporates a nonvolatile random access memory (NVRAM) module that stores the backplane serial number for identification and revision control purposes. The contents of the NVRAM module are accessible from any line card slot.

Cable-Management System

The cable-management system provides an orderly and convenient way for you to manage the network interface cables running to and from the receive and transmit ports of installed line cards.

Consisting of a cable-management tray and a vertical cable-management bracket (one bracket for each installed line card), the cable-management system (see ) secures the network interface cables neatly in place. The cable management system helps to optimize optical cable performance by eliminating any kinks or sharp bends in the cables. Extreme curvatures in optical cables tend to degrade their performance.

The elements of the cable-management system are shown in and described briefly in the following sections:

Cable-management tray—This tray is attached to the router enclosure above the upper card cage.

The cable management tray enables you to route the line card interface cables to or from the system through the left side of the tray, keeping the cables organized, out of the way, and free of kinks or sharp bends.

You direct the cables down to the individual ports on each line card, gauging cable length appropriately to minimize slack in the cable before connecting it to a given port.

Figure 1-3 Cable-Management System

Vertical cable-management bracket (one per line card)—This bracket is attached to a line card by means of captive installation screws at the top and bottom of the bracket.

Once an interface cable is connected to its intended line card port, you loop the cable through the cable keeper clip nearest the port of connection and seat the cable in the bottom of the bracket raceway.

Thus, the vertical cable-management bracket enables you to neatly "dress" all the interface cables in place as you connect them to the individual line card ports.

Later, when you remove or replace a line card, you need only disconnect the cables from the individual line card ports (leaving the cables intact within the vertical cable-management bracket) and detach the bracket from the line card to be replaced.

When you install the new line card, you merely reattach the vertical cable-management bracket to the new line card and reconnect the interface cables to the appropriate line card port(s).

Card Cage Fan Tray

The card cage fan tray is located in the lower card cage behind the air filter assembly (see ). This fan tray maintains the operating temperature of the router's electronic circuitry within an acceptable range.

Designed for simplicity, the card cage fan tray incorporates six fans mounted on a sheet metal carrier. The assembly also contains associated wiring and a connector in the back of the unit that enables it to draw operating power through the backplane from a DC-DC converter on the CSC.

Guide rails in the sides of the lower card cage facilitate insertion and removal of the fan tray assembly, which is secured in place by means of a captive installation screw on each side of the metal carrier.

Under normal operating conditions, the variable-speed fans in the card cage fan tray operate at a reduced rate to

Conserve power

Reduce noise

Minimize fan wear

If an overtemperature condition or a fan failure is detected within the router, the master MBus module on the RP directs the MBus module on the clock and scheduler card (CSC) to increase the operating voltage being delivered to the fan tray, causing the card cage fans to run at "maximum" speed. This increases the volume of cooling air flowing through the router.

If the increased fan speed does not alleviate the overtemperature condition in the affected board, the MBus module on the board shuts down the board's power supply, taking the board offline to protect it from thermal damage.

The MBus facility of the Cisco 12008 router is described in greater detail in the section entitled "Cisco 12008 Environmental Monitoring Facility" on page 75.

Power Supply Fan Tray

The power supply fan tray is in the bottom of the power supply bays (see ). This fan tray maintains the temperature of the installed power supply(ies) within an acceptable range.

Also designed for simplicity, the power supply fan tray incorporates four fans mounted on a sheet metal carrier. The fan tray assembly contains associated wiring and a connector in the back of the unit that enables it to draw operating power through the backplane from a DC-DC converter on the CSC.

A captive installation screw mounted on the fan tray faceplate and guide rails in the sides of the power supply bay facilitate insertion and removal of the unit. Once the unit is inserted, you secure it in place by tightening the captive installation screw clockwise.

Similar to the card cage fan tray, the power supply fan tray is closely tied to the router's overall environmental monitoring system. If an overheating condition or a fan failure is detected within the router, the voltage being delivered to the power supply fans by the CSC is also increased, thereby causing the power supply fans to run at "maximum speed" to increase the volume of cooling passing through the power supply bays.

AC-Input and DC-Input Power Supplies

The Cisco 12008 router can be configured to operate with AC source power or DC source power. You can install one or two AC-input power supplies or one or two DC-input power supplies in the power supply bays located in the right side of the router enclosure (see ).

A single power supply of either type is the standard router configuration. In such a configuration, it is recommended that you install the power supply in the lower bay.

You can install a second (optional and redundant) power supply of the same type for backup purposes.


Caution   
A vacant power supply bay must be covered with a blank filler panel to ensure proper flow of cooling air through the power supply bays and to satisfy EMI compliance requirements.


Note   You cannot use an AC-input power supply in conjunction with a DC-input power supply. Installed power supplies must always be of the same type. Furthermore, you should not install two power supplies of either type unless you intend to actively use both units. In other words, you should not power the router with a single power supply while using the other bay to temporarily or indefinitely "store" an inert unit. Doing so will disrupt the normal flow of cooling air through the router enclosure.


shows an AC-input power supply; shows a DC-input power supply.

Figure 1-4 AC-Input Power Supply

Figure 1-5 DC-Input Power Supply

Operating Modes of the Power Supplies

The AC-input and DC-input power supplies operate in either of two modes:

Standalone mode—In this configuration, only one power supply is installed in one of the two available power supply bays. To remove or replace a single power supply, you must first power down the system.

Redundant (1+1) mode—In this configuration, two power supplies are installed in the power supply bays, sharing the load current to provide required DC operating voltages to the backplane. If one of the units fails, the surviving power supply takes over to maintain normal system operations.

The online insertion and removal (OIR) capability of the router enables you to add or remove a redundant power supply without introducing noise in the DC operating voltages being supplied to the backplane.

Features of the Power Supplies

The AC-input and DC-input power supplies incorporate the following features:

Onboard maintenance bus (MBus) module—The MBus module on the power supply is a microprocessor-based subassembly that links the power supply to the router's environmental monitoring system.

The environmental monitoring system includes identical MBus modules on all of the router circuit boards, including the RP. This system enables you to perform router functions and to respond to alarm conditions (such as overtemperature or overvoltage conditions).

An alarm condition in the router causes the MBus module on the CSC to illuminate an appropriate LED on the card faceplate, providing a visible notification of the alarm condition.

Blind mating connector at the back of the unit—Supplies DC operating voltages to the backplane for distribution to the router's electronic and electrical components.

OIR capability—Enables a second AC-input power supply to be installed in or removed from the router without disrupting normal system operations.

Temperature sensor—Measures the ambient air temperature of the power supply.

Characteristics of the Power Supplies

The AC-input and the DC-input power supplies have the following characteristics:

Width of power supply body—3.5 inches (8.97 cm)

Width of power supply faceplate—4.0 inches (10.26 cm)

Height—10 inches (25.64 cm)

Depth—17.6 inches (45.13 cm)

Weight (AC-input power supply)—17 lb (7.73 kg)

Weight (DC-input power supply)—14 lb (6.36 kg)

Power factor corrector (PFC)—Applicable only to the AC-input power supply, the PFC enables the power supply to accept source AC voltages with the following characteristics: voltages ranging from 180 to 264 VAC, single phase, 47 to 63 Hz.

AC-Input Power Supply Faceplate

This section describes the functional elements built into the faceplate of the AC-input power supply (see ).

Figure 1-6 AC-Input Power Supply Faceplate

Rotary Power Switch

The rotary power switch on the power supply faceplate (see ) applies a source AC voltage to the power supply. This switch also actuates an onboard circuit breaker and a latching mechanism that prevents the power supply from being inserted into or removed from the power supply bay when the switch is in the ON (1) position.

When you rotate the rotary power switch 90 degrees to the ON position, the following DC operating voltages are supplied to the backplane:

+5.2 VDC

-48 VDC

Source AC Input Connector

The source AC receptacle on the power supply faceplate (see ) enables an external AC power source to be connected to the power supply. This connector is equipped with a latch that prevents accidental or unintended removal of the AC power cord.

The power specifications for the AC-input power supplies, as well as the source AC power cables available for use with the Cisco 12008 router, are described in Chapter 2 in the section entitled "AC-Powered Systems."

AC-Input Power Supply LEDs

The AC-input power supply faceplate incorporates two LEDs (see ) that provide the following status indications:

AC INPUT OK—When the rotary power switch is turned ON, this green LED goes on, indicating that source AC power has been applied and that it is within the specified operating range. If this LED does not go on when the rotary power switch is turned ON, it indicates that source AC power is not within the specified operating range or that the LED is faulty.

OUTPUT FAIL—When the rotary power switch is turned on, this LED goes on momentarily; it should then go off and remain so. If it does not go off, it indicates that the +5.2 VDC or -48 VDC being supplied to the backplane is not within tolerance.

DC-Input Power Supply Faceplate

This section describes the functional elements built into the faceplate of the DC-input power supply (see ).

Figure 1-7 DC-Input Power Supply Faceplate

Rotary Power Switch

The rotary power switch on the DC-input power supply performs the same functions as those described in the section entitled "Rotary Power Switch" on page 18 for the AC-input power supply.

Circuit Breaker Alarm Terminal Block

The onboard power supply circuit breaker actuated by the rotary power switch on the DC-input power supply incorporates an auxiliary switch that is mechanically linked to (but electrically isolated from) the power supply circuit breaker.

When the power supply circuit breaker is tripped by an overcurrent condition in the power supply, this auxiliary switch moves in unison, sending a signal to the circuit breaker alarm terminal block on the power supply faceplate (see ).

To remotely sense when the power supply circuit breaker has been tripped during an overcurrent condition, you can attach an external alarm-monitoring facility to the alarm terminal block. When the power supply circuit breaker is tripped, power is no longer delivered to the backplane and the router ceases to operate. Hence, if you have attached an external alarm monitoring facility to the alarm terminal block, site personnel can be instantly alerted to this serious fault condition.

Typically, an external alarm-monitoring system incorporates a light panel (visible alarm) or a klaxon (audible alarm) as the means for alerting site personnel to an alarm condition.

To reset the alarm contacts on the alarm terminal block, you must turn the rotary power switch on the power supply OFF and then ON again, much as you would reset any circuit breaker.


Note   Any time you manually actuate the rotary power switch, such as when powering down the router, the contacts on the alarm terminal block remain unaffected. Hence, activation of the contacts on the alarm terminal block occurs only during a power supply overcurrent condition. In other words, these contacts are used to provide an immediate, overt indication of a power supply fault condition; they are not used to merely indicate that a circuit breaker has been turned off manually.


The three contacts on the alarm terminal block are labeled as follows:

COM (Common)—This contact is common to both the Normally Open (NO) and the Normally Closed (NC) contacts.

NO (Normally Open)—These contacts on the alarm terminal block are open as long as no overcurrent condition is detected in the power supply. When the power supply circuit breaker is tripped during an overcurrent condition, these contacts are closed.

NC (Normally Closed)—These contacts on the alarm terminal block are closed as long as no overcurrent condition is detected in the power supply. When the power supply circuit breaker is tripped during an overcurrent condition, these contacts are open.

summarizes the status of the contacts on the alarm terminal block during an overcurrent condition in the power supply.

Table 1-1 Circuit Breaker Status Indicated by the Alarm Terminal Block

Circuit Breaker Position
NC Contact
NO Contact

OFF (tripped)

Open

Closed

ON

Closed

Open


If you decide to use an external alarm-monitoring facility in conjunction with the alarm terminal block, note that the contacts on the alarm terminal block have a rating of 60 VDC at 1A maximum.

Source DC Input Connectors

The faceplate of the DC-input power supply incorporates three sets of terminals for connecting source DC power to the power supply (see ). From top to bottom, these terminals are identified as follows:

Ground

+ (positive)

- (negative)

The power specifications for the DC-input power supplies, as well as the specifications of the source DC power cables for use with the Cisco 12008 router, are presented in the section entitled "DC-Powered Systems" on page 16 in Chapter 2.

DC-Input Power Supply LEDs

The DC-input power supply faceplate incorporates two LEDs (see ) that provide the following status indications:

INPUT OK—When the rotary power switch is turned ON, this green LED goes on immediately, indicating that source DC power is applied and that it is within the specified operating range (-40.5 VDC to -75 VDC). If this LED does not go on when the rotary power switch is turned ON, the source DC power being applied to the power supply is not within the normal operating range or the LED is faulty.

OUTPUT FAIL—When the rotary power switch is turned on, this LED goes on momentarily; it should then go off and remain so. If it does not go off, it indicates that the +5.2 VDC or -48 VDC being supplied to the backplane is not within tolerance.

Upper Card Cage and Associated Components

The upper card cage (see ) contains ten slots that accommodate the following types of cards in the quantities indicated:

One Route Processor (RP)—A RP is a standard and required router component; the RP must be present and operational at all times. It is recommended that you install the RP in the left-most slot (slot 0) in the upper card cage.

Either one or two clock and scheduler cards (CSCs)—One CSC is a standard and required router component; one CSC must be present and operational in the router at all times. For redundancy, you can install a second CSC for use as a backup.

Two dedicated slots in the middle of the upper card cage (CSC0 and CSC1) are reserved for the CSCs. Because the backplane connector of a CSC differs significantly from all other card types, you cannot install a CSC in any other slot.

Cisco 12000 series line cards—From one to seven line cards of different types can be installed in the line card slots in the upper cage (slots 0 through 3 and slots 4 through 7).

Although you can install a line card in slot 0, the recommended convention is for the RP to occupy this slot.

Figure 1-8 Upper Card Cage of the Cisco 12008 Router

A minimally configured Cisco 12008 contains the following cards in the upper card cage:

One RP

One CSC

One Cisco 12000 series line card of any type

A Cisco 12008 that is configured for full redundancy contains the following cards in the upper card cage:

Two RPs

Two CSCs

As many as six Cisco 12000 series line cards of any type and any combination

The following sections briefly describe the cards that you can use to populate the upper card cage.

Gigabit Route Processor

Each Cisco 12008 GSR has one main system (or route) processor. The route processor (RP) processes the network routing protocols and distributes updates to the Cisco Express Forwarding (CEF) tables on the line cards. The RP also performs general maintenance functions, such as diagnostics, console support, and line card monitoring.

Two types of RPs are available for the Cisco 12008 GSR:

Gigabit Route Processor (GRP)

Performance Route Processor (PRP)

When not explicitly specified, this document uses the term route processor (RP) to indicate either the GRP or the PRP.


Note   If you install a second, redundant RP, it must be of the same type as the primary RP.


This section describes the GRP and includes the following information:

Memory components

System status LEDs

Soft reset switch

Personal Computer Memory Card Industry Association (PCMCIA) slots, which are used to transmit data to or from Flash memory cards

Asynchronous serial ports

Ethernet port

If you have a PRP, see the Performance Route Processor section.

The faceplate of the GRP is shown in .

Figure 1-9 GRP Faceplate (Horizontal Orientation Shown)

It is recommended that you install the GRP in the left-most slot (slot 0) in the upper card cage. However, you need not abide by this recommendation. You can install the GRP in any upper card cage slot, except for the two slots in the middle in the upper card cage (CSC0 and CSC1), which are reserved for the CSCs.

The GRP performs the following functions:

Downloading the Cisco IOS software to all of the installed line cards at power up

Providing a console (terminal) port for router configuration

Providing an auxiliary port for other external equipment (such as modems)

Providing an IEEE 802.3, 10/100-megabits-per-second (Mbps) Ethernet port for Telnet functionality

Running routing protocols

Building and distributing routing tables to line cards

Providing general system maintenance functions

The GRP communicates with the line cards either through the switch fabric or through a maintenance bus (MBus). The switch fabric connection is the main data path for routing table distribution as well as for packets that are sent between the line cards and the GRP. The MBus connection allows the GRP to download a system bootstrap image, collect or load diagnostic information, and perform general, internal system maintenance operations. The GRP plugs into any slot in the upper card cage in the Cisco 12008 except the rightmost slot, which is reserved for the alarm card.

The GRP contains the following components:

IDT R5000 Reduced Instruction Set Computing (RISC) processor used for the CPU. The CPU runs at an external bus clock speed of 100 MHz and an internal clock speed of 200 MHz.

Up to 256 megabytes (MB) of parity-protected, extended data output (EDO) dynamic random-access memory (DRAM) on two, 60-nanosecond (ns), dual in-line memory modules (DIMMs); 64 MB of DRAM is the minimum shipping configuration.

512 kilobytes (KB) of static random-access memory (SRAM) for secondary CPU cache memory functions (SRAM is not user configurable or field upgradeable).

512 KB of NVRAM (NVRAM is not user configurable or field upgradeable).

Most of the additional memory components used by the system, including onboard Flash memory (8-MB) and up to two PCMCIA-based Flash memory cards. The default GRP PCMCIA Flash memory is 20 megabytes (MB).

Air-temperature sensors for environmental monitoring.

The Cisco IOS software images that run the Cisco 12008 reside in Flash memory, which is located on the GRP in the form of a single in-line memory module (SIMM), and on up to two Personal Computer Memory Card International Association (PCMCIA) cards (called Flash memory cards) that insert in the two PCMCIA slots (slot 0 and slot 1) on the front of the GRP. (See .) Storing the Cisco IOS images in Flash memory enables you to download and boot from upgraded Cisco IOS images remotely or from software images resident in GRP Flash memory.


Note   EIA/TIA-232 was previously known as recommended standard RS-232 before its acceptance as a standard by the Electronic Industries Association (EIA) and the Telecommunications Industry Association (TIA).


The Cisco 12008 supports downloadable system software for most Cisco IOS software upgrades, enabling you to remotely download, store, and boot from a new Cisco IOS software image.

GRP Memory Components

lists the memory components on the GRP. shows the location of the two DRAM SIMMs and the Flash SIMM on the GRP.

Table 1-2 GRP Memory Components

Memory Type
Memory Size
Quantity
Description

DRAM

641 to 256 MB

1 or 2

64- or 128-MB DIMMs (based on DRAM required) for main Cisco IOS software functions

SRAM

512 KB (fixed)2

 

SRAM for secondary CPU cache memory functions

NVRAM

512 KB (fixed)3

 

MVRAM for the system configuration file

Flash memory SIMM4

8 MB

1

Contains Cisco IOS software images and other user-defined files on the GRP

Flash memory (card)

20 MB5

1 or 2

Contains Cisco IOS software images and other user-defined files on up to two PCMCIA-based Flash memory cards6

Flash boot ROM

512 KB

1

Flash EPROM for the ROM monitor program boot image

1 64 MB of DRAM is the default DRAM configuration for the GRP.

2 SRAM is not user configurable or field upgradeable.

3 NVRAM is not user configurable or field upgradeable.

4 The SIMM socket is wired according to Cisco's own design and does not accept industry-standard 80-pin Flash SIMMs.

5 20-MB Flash memory card is the default shipping configuration for the Cisco 12008.

6 A Type 1 or Type 2 PCMCIA card can be used in either PCMCIA slot.


Figure 1-10 Locations of GRP Memory

DRAM

The extended data output (EDO) dynamic random-access memory (DRAM) on the GRP stores routing tables, protocols, and network accounting applications, and runs the Cisco IOS software. The standard (default) GRP DRAM configuration is 64 megabytes (MB) of EDO DRAM, which you can increase up to 256 MB through DRAM upgrades. The Cisco IOS software runs from within GRP DRAM.

Two DRAM DIMM sockets are incorporated into the GRP, as shown in . These sockets, labeled U39 (P4 DRAM bank 1) and U42 (P4 DRAM bank 2), enable you to configure DRAM in increments ranging from 64 MB to 256 MB. lists the available upgrade configurations for DRAM on the GRP.

Table 1-3 DRAM Configurations

Total DRAM
Product Numbers
DRAM Sockets
Number of DIMMs

64 MB1

MEM-GRP/LC-64(=)

U39 (bank 1)

1 64-MB DIMM

128 MB

MEM-GRP/LC-64(=)

U39 (bank 1) and U42 (bank 2)

2 64-MB DIMMs

128 MB

MEM-GRP/LC-128(=)

U39 (bank 1)

1 128-MB DIMM

256 MB

MEM-GRP/LC-256(=)

U39 (bank 1) and U42 (bank 2)

2 128-MB DIMMs

1 64 MB is the standard (default) DRAM configuration for the GRP.



Caution   
To prevent memory problems, DRAM DIMMs must be 3.3 V, 60-nanosecond (ns) devices. Do not attempt to install memory devices in the DIMM sockets that do not meet these requirements.

SRAM

SRAM provides secondary CPU cache memory. The standard GRP configuration is 512 KB. Its principle function is to act as a staging area for routing tables update information to and from the line cards. SRAM is not user configurable or field-upgradeable.

NVRAM

The system configuration, software configuration register settings, and environmental monitoring logs are contained in the 512-KB NVRAM, which is backed up with built-in lithium batteries that retain the contents for a minimum of five years. NVRAM is not user configurable or field-upgradeable.


Caution   
Before you replace the GRP in the system, back up the running configuration to a Trivial File Transfer Protocol (TFTP) file server or an installed Flash memory card so you can retrieve it later. If the configuration is not saved, the entire configuration will be lost—inside the NVRAM on the removed GRP—and you will have to reenter the entire configuration manually. This procedure is not necessary if you are temporarily removing a GRP; lithium batteries retain the configuration in memory until you replace the GRP in the system.

Flash Memory

Both the onboard and PCMCIA card-based Flash memory allow you to remotely load and store multiple Cisco IOS software and microcode images. You can download a new image over the network or from a local server and then add the new image to Flash memory or replace the existing files. You can then boot the routers either manually or automatically from any of the stored images. Flash memory also functions as a TFTP server to allow other servers to boot remotely from stored images or to copy them into their own Flash memory.

System Status LEDs

This section describes the two types of system status LEDs used on the GRP: the LED indicators and the alphanumeric LED displays.

The GRP has the following eight LED indicators:

Two PCMCIA activity LEDs (one per PCMCIA slot): these LEDs light when the slot is accessed. The LEDs receive power from the switched slot voltage.

Four RJ-45 Ethernet port LEDs: these LEDs are used in conjunction with the RJ-45 Ethernet connector. When the MII Ethernet port is in use, the LEDs are disabled. The LEDs indicate link activity, collision detection, data transmission, and data reception.

Two RJ-45 or MII Ethernet port select LEDs: these LEDs, when on, identify which one of the two Ethernet connections you selected. When the RJ-45 port is selected, its LED is on and the MII LED is off. When the MII port is selected, its LED is on and the RJ-45 LED is off.

The alphanumeric displays are organized as two rows of four characters each. The displays' content is controlled by the MBus module software. The displays' content is controlled by the GRP's MBus module software. Both rows of the display are powered by the MBus module.

These alphanumeric displays provide information about the following:

System status messages that are displayed during the boot process

System status messages that are displayed after the boot process is complete

During the boot process, the alphanumeric LED displays are controlled directly by the MBus. After the boot process, they are controlled by the Cisco IOS software (via the MBus), and display messages designated by the Cisco IOS software.

The following levels of system operation are displayed:

Status of the GRP

System error messages

User-defined status/error messages


Note   A complete, descriptive list of all system and error messages is located in the Cisco IOS System Error Messages publications.


Soft Reset Switch

A soft reset switch is provided on the GRP faceplate to enable you to reset the software running on the R5000 RISC processor of the GRP. You access this switch through a small aperture in the GRP faceplate. To activate the switch, you can press a ball-point pen or similar pointed instrument into the opening.


Caution   
To prevent system problems or loss of data, use the soft reset switch only at the advice of Cisco service personnel.

PCMCIA Slots

The GRP has two PCMCIA slots available. Either slot can support a Flash memory card or an input/output (I/O) device as long as the device requires only +5 VDC. The GRP supports Type 1 and Type 2 devices; it does not support +3.3 VDC PCMCIA devices. Each PCMCIA slot has an ejector button for ejecting a PCMCIA card from the slot.

Asynchronous Serial Ports

Two asynchronous serial ports are provided on the GRP faceplate—a console port and an auxiliary port. These ports enable you to connect external devices that you can use to monitor and manage the system.

Console port—The console port is an Electronics Industries Association/Telecommunications Industry Association (EIA/TIA)-232 female receptacle that provides a data circuit-terminating equipment (DCE) interface for connecting a console terminal.

Auxiliary port—The auxiliary port is an EIA/TIA-232 male plug that provides a data terminal equipment (DTE) interface. This auxiliary port supports flow control and is often used to connect a modem, a channel service unit (CSU), or other optional equipment for Telnet management.

Ethernet Port

The GRP has one Ethernet port that you can access using either of the following connection types:

RJ-45 receptacle: an 8-pin, media dependent interface (MDI) that supports an IEEE 802.3 10BaseT (10 Mbps) or an IEEE 802.3u 100BaseTX (100 Mbps) Ethernet connection.

MII receptacle: a 40-pin, media independent interface (MII) that provides additional flexibility for making Ethernet connections. The pinout of this standard 40-pin receptacle is defined by the IEEE 802.3u standard.


Note   The RJ-45 and MII receptacles on the GRP faceplate represent two physical connection options for one Ethernet interface; therefore, you can use either the MDI RJ-45 connection or the MII connection, but not both simultaneously.


Performance Route Processor

Each Cisco 12012 GSR has one main system (or route) processor. The route processor (RP) processes the network routing protocols and distributes updates to the Cisco Express Forwarding (CEF) tables on the line cards. The RP also performs general maintenance functions, such as diagnostics, console support, and line card monitoring.

Two types of RPs are available for the Cisco 12012 GSR:

Gigabit Route Processor (GRP)

Performance Route Processor (PRP)

When not explicitly specified, this document uses the term route processor (RP) to indicate either the GRP or the PRP.


Note   If you install a second, redundant RP, it must be of the same type as the primary RP.


The section describes the Performance Route Processor (PRP) and includes the following information:

PRP Memory Components

System Status LEDs

Soft Reset Switch

PCMCIA Slots

Asynchronous Serial Ports

Ethernet Port

If you have a GRP, see the Gigabit Route Processor section.

shows the front panel view of the PRP.

Figure 1-11 Performance Route Processor (Front Panel View, Horizontal Orientation Shown)

The PRP is available as Product Number PRP-1=, which includes one PRP with 512 MB of synchronous dynamic random-access memory (SDRAM) and one 64-MB advanced technology attachment (ATA) Flash disk.

The primary functions of the PRP are as follows:

Downloading the Cisco IOS software to all of the installed line cards at power up

Providing a console (terminal) port for router configuration

Providing an auxiliary port for other external equipment (such as modems)

Providing two IEEE 802.3, 10/100-megabits-per-second (Mbps) Ethernet ports for Telnet functionality

Running routing protocols

Building and distributing routing tables to line cards

Providing general system maintenance functions

Communicating with line cards either through the switch fabric or through the maintenance bus (MBus)

The MBus connection allows the PRP to download a system bootstrap image, collect or load diagnostic information, and perform general, internal system maintenance operations. The switch fabric connection is the main data path for routing table distribution as well as for packets that are sent between line cards and the PRP.

The PRP contains the following components:

Motorola PowerPC 7450 central processing unit (CPU). The CPU runs at an external bus clock speed of 133 MHz and an internal clock speed of 667 MHz.

Up to 2 GB of SDRAM on two PC133-compliant, dual in-line memory modules (DIMMs). 512 MB of SDRAM is the default shipping configuration. SDRAM is field replaceable.

Two MB of SRAM for secondary CPU cache memory functions. SRAM is not user configurable or field replaceable.

Two MB of NVRAM. NVRAM is not user configurable or field replaceable.

Additional memory components used by the system, including onboard Flash memory and up to two Flash memory cards.

Air-temperature sensors for environmental monitoring.

The Cisco IOS software images that run the Cisco 12000 series Internet Router system are stored in Flash memory. Two types of Flash memory ship with the PRP:

1 Onboard Flash memory — Ships as a single in-line memory module (SIMM). This Flash memory contains the Cisco IOS boot image (bootflash) and is not field replaceable.

2 Flash disk— The PRP ships with a Flash disk that can be installed in either Flash disk slot. (See .) The Flash disk contains the Cisco IOS software image.

Storing the Cisco IOS images in Flash memory enables you to download and boot from upgraded Cisco IOS software images remotely, or from software images that reside in PRP Flash memory.

Cisco 12000 series Internet Routers support downloadable system software for most Cisco IOS software upgrades. This enables you to remotely download, store, and boot from a new Cisco IOS software image. The Cisco IOS software runs from within the PRPs SDRAM.

shows the locations of the various hardware components on the PRP.

Figure 1-12 PRP (Horizontal Orientation)

1

Backplane connector

6

Ethernet ports

2

Flash SIMM (Socket number P3)

7

Auxiliary port

3

SDRAM DIMMs
Bank 1 - Socket number U15
Bank 2 - Socket number U18

8

Console port

4

Ejector lever

9

Handle

5

Flash disk slots (covered)

10

Display LEDs


PRP Memory Components

lists the memory components on the PRP.

Table 1-4

Type
Size
Quantity
Description

SDRAM1

512 MB, 1 GB, or 2 GB

1 or 2

512-MB and 1-GB DIMMs (based on desired SDRAM configuration) for main Cisco IOS software functions

SRAM2

2 MB (fixed)

Secondary CPU cache memory functions

NVRAM3

2 MB (fixed)

1

System configuration files, register settings, and logs

Flash memory

64 MB SIMM4

1

Cisco IOS boot image (bootflash), crash information, and other user-defined files

 

Flash disks5

1 or 2

Cisco IOS software images, system configuration files, and other user-defined files on up to two Flash disks

Flash boot ROM

512 KB

1

Flash EPROM for the ROM monitor program boot image

1 Default SDRAM configuration is 512 MB. Bank 1 (U15) must be populated first. You can use one or both banks to
configure SDRAM combinations of 512 MB, 1 GB, or 2 GB. 1.5-GB configurations are not supported.

2 SRAM is not user configurable or field replaceable.

3 NVRAM is not user configurable or field replaceable.

4 Flash memory SIMM is not user configurable or field replaceable.

5 ATA Flash disks and Type I and Type II linear Flash memory cards are supported. See the
"Flash Memory" for Flash disk information.


PRP Memory Components


Note   If a single DIMM module is installed, it must be placed in bank 1 (U15).


SDRAM

SDRAM stores routing tables, protocols, and network accounting applications, and runs the Cisco IOS software. The default PRP configuration includes 512 MB of error checking and correction (ECC) SDRAM. DIMM upgrades of 512 MB and 1 GB are available. You cannot mix memory sizes. If two DIMMS are installed, they must be the same memory size.


Caution    Cisco Systems strongly recommends that you use only Cisco-approved memory. To prevent memory problems, SDRAM DIMMs must be +3.3VDC, PC133-compliant devices. Do not attempt to install other devices in the DIMM sockets.

SRAM

SRAM provides 2 MB of parity-protected, secondary CPU cache memory. Its principal function is to act as a staging area for routing table updates and for information sent to and received from line cards. SRAM is not user configurable and cannot be upgraded in the field.

NVRAM

NVRAM provides 2 MB of memory for system configuration files, software configuration register settings, and environmental monitoring logs. This information is backed up with built-in lithium batteries that retain the contents for a minimum of 5 years. NVRAM is not user configurable and cannot be upgraded in the field.

Flash Memory

Flash memory allows you to remotely load and store multiple Cisco IOS software and microcode images. You can download a new image over the network or from a local server and then add the new image to Flash memory or replace the existing files. You then can boot the routers either manually or automatically from any of the stored images.

Flash memory also functions as a Trivial File Transfer Protocol (TFTP) server to allow other servers to boot remotely from stored images or to copy them into their own Flash memory. The onboard Flash memory (called bootflash) contains the Cisco IOS boot image, and the Flash disk contains the Cisco IOS software image. A 64-MB ATA Flash disk ships by default with the PRP. lists the supported Flash disk sizes and their Cisco product numbers.

Table 1-5

Flash Disk Size 1
Product Number

64 MB2

MEM-12KRP-FD64=

128 MB

MEM-12KRP-FD128=

1 GB

MEM-12KRP-FD1G=

1 Standard Type 1 and Type 2 linear Flash memory cards also are supported, although they may not have the capacity to meet the requirements of your configuration.

2 64-MB ATA Flash disk is the default shipping configuration.


Supported Flash Disk Sizes and Product Numbers

System Status LEDs

The sections describes the two types of system status LEDs used on the PRP: LED indicators and alphanumeric LED displays.

The device or port activity indicators consist of the following functional groups:

Two Flash disk activity LEDs (labeled SLOT-0 and SLOT-1)—1 LED per Flash disk slot: these go on when the slot is accessed.

Four RJ-45 Ethernet port LEDs (labeled LINK, EN, TX, and RX): used in conjunction with each of the RJ-45 Ethernet connectors. Each connector includes a set of 4 LEDs that indicate link activity (LINK), port enabled (EN), data transmission (TX), and data reception (RX).

Two Ethernet connection LEDs (labeled PRIMARY): these two LEDs, when on, identify which of the two Ethernet connections is selected. Since both ports are supported on the PRP, the LED on port ETH0 is always on. The ETH1 LED goes on when it is selected.

The alphanumeric display LEDs are organized as two rows of four characters each and are located at one end of the card. These LEDs provide system status and error messages that are displayed during and after the boot process. The boot process and the content displayed are controlled by the PRPs MBus module software.

At the end of the boot process, the LEDs are controlled by the Cisco IOS software (via the MBus), and the content displayed is designated by the Cisco IOS software.

The alphanumeric display LEDs provide information about the following:

Status of the PRP

System error messages

User-defined status and error messages


Note   A complete, descriptive list of all system and error messages is located in the Cisco IOS System Error Messages publications.


Soft Reset Switch

The soft reset switch causes a nonmaskable interrupt (NMI) and places the PRP in ROM monitor mode. When the PRP enters ROM monitor mode, its behavior depends on the setting of the PRP software configuration register. (For more information on the software configuration register, refer to the Configuring the Software Configuration Register section in Chapter 4) For example, when the boot field of the software configuration register is set to 0x0, and you press the NMI switch, the PRP remains at the ROM monitor prompt (rommon>) and waits for a user command to boot the system manually. But if the boot field is set to 0x1, the system automatically boots the first IOS image found in the onboard Flash memory SIMM on the PRP.


Caution   
The soft reset (NMI) switch is not a mechanism for resetting the PRP and reloading the IOS image. It is intended for software development use. To prevent system problems or loss of data, use the soft reset switch only on the advice of Cisco service personnel.

Access to the soft reset switch is through a small opening in the PRP faceplate. To press the switch, you must insert a paper clip or similar small pointed object into the opening.

Flash Disk Slots

The PRP includes two Flash disk (PCMCIA) slots. Either slot can support an ATA Flash disk or a Type 1 or Type 2 linear Flash memory card. The PRP ships by default with one 64-MB ATA Flash disk.


Note   The PRP only supports +5VDC Flash disk devices. It does not support +3.3VDC PCMCIA devices.


All combinations of different Flash devices are supported by the PRP. You can use ATA Flash disks, Type 1 or Type 2 linear Flash memory cards, or a combination of the two. Each Flash disk slot has an ejector button for ejecting a card from the slot.


Note   Type 1 and Type 2 linear Flash memory cards may not have the capacity to meet the requirements of your configuration.


Asynchronous Serial Ports

The PRP has two asynchronous serial ports, the console and auxiliary ports. These allow you to connect external serial devices to monitor and manage the system. Both ports use RJ-45 receptacles.

The console port provides a data circuit-terminating equipment (DCE) interface for connecting a console terminal. The auxiliary port provides a data terminal equipment (DTE) interface and supports flow control. It is often used to connect a modem, a channel service unit (CSU), or other optional equipment for Telnet management.

Ethernet Ports

The PRP includes two Ethernet ports, both using an 8-pin RJ-45 receptacle for either IEEE 802.3 10BASE-T (10 Mbps) or IEEE 802.3u 100BASE-TX (100 Mbps) connections.


Note   The transmission speed of the Ethernet ports is auto-sensing by default and is user configurable.


Switch Fabric of the Cisco 12008

The heart of the Cisco 12008 is the switch fabric circuitry, which provides synchronized gigabit speed interconnections between the line cards and the RP. The switch fabric circuitry for the router is incorporated into two cards:

Clock and scheduler card (CSC)—One CSC installed in the upper card cage is a standard (required) router component. The CSC represents one plane of switch fabric in the router. This card is described in greater detail in the section entitled "Clock and Scheduler Card."

Switch fabric card (SFC)—You can install a set of three optional SFCs in the lower card cage to increase its switching (data-handling) capacity). Each SFC card represents one plane of switch fabric in the router. This card is described in greater detail in the section entitled "Switch Fabric Cards."

To achieve a fully redundant switch fabric with a switching capacity of 40 Gbps, you can install two CSCs and three SFCs in the router; the second CSC provides redundancy of CSC functions, as well as redundant switch fabric in the event of CSC or SFC failure.

Each CSC or SFC supports an OC-12 switching rate for the router (622 Mbps). By adding the set of three optional SFC cards, you can increase the switching capacity of the router to an OC-48 rate (2.4 Gbps).

lists the switch fabric bandwidth and the switch card configurations needed to support an OC-12 switching rate or an OC-48 switching rate.

Table 1-6 Switch Fabric Configurations

Switch Fabric Bandwidth
Number of CSCs
Number of SFCs
Planes of Switch Fabric

OC-12 nonredundant

11

0

1

OC-12 redundant

2

0

2

OC-48 nonredundant

1

3

4

OC-48 redundant

2

3

5

1 A CSC is a required router component.


A minimally configured router (one with a single CSC and no SFCs) supports an OC-12 data rate, but provides no redundancy of CSC functions. Adding a second CSC to a system, as well as the three optional SFCs, has the following effects:

Increases the router's bandwidth from an OC-12 rate to an OC-48 rate.

Increases the number of planes of switch fabric available to the router from one to five (with the fifth serving as a redundant plane in the event of failure of a CSC or SFC.

Provides full redundancy of CSC functions, such as the following:

System clocking

Resource allocation

Scheduling

Provides full redundancy in the router's fan power and alarm functions.

Clock and Scheduler Card

The CSC is a multi-function circuit board that can be installed in one or both of two reserved slots (CSC0 and CSC1) in the middle of the upper card cage (see ). The standard router configuration requires one CSC in either slot CSC0 or slot CSC1. If you configure your router with a single CSC, it is recommended that you install it in CSC1.

Each CSC is mounted on its own card carrier and incorporates an onboard power supply that takes the -48 VDC supplied by the backplane and converts it into the 3.3 VDC operating voltage required by the card's electronics.

As a multi-function board, the CSC provides the following system services:

Provides one plane of switch fabric for the router (see the section below entitled "Switch Fabric in the Cisco 12008").

Serves as a switch fabric controller card for the router (see the section below entitled "Switch Fabric Controller Functions of the CSC").

Serves as an alarm monitoring facility for the router (see the section below entitled "Housekeeping and Alarm Monitoring Functions of the CSC").

Provides onboard power for its own electronic circuitry, as well as power and control functions for the fan trays (see the section below entitled "Board Power and Fan Tray Power Functions of the CSC").

These functions are described in the following sections.

Switch Fabric in the Cisco 12008

A switch plane in the router consists of one OC-12-rate crossbar in the backplane that enables each line card slot in the router to be connected logically to every other line card slot. Line cards installed in any combination of slots in the upper card cage can communicate with each other by means of the router's switch fabric.

The switch fabric of the router constitutes the totality of the possible data paths that can be established through the router. The magnitude of the router's switch fabric (and, hence, its data- carrying capacity) is related directly to the number of switch planes that are made available to the router for data-handling purposes. By installing a second CSC and/or the optional set of three SFCs in the router, you can increase the number of switch planes present in the router, thereby increasing the magnitude of the router's overall switch fabric.

outlines the possible configurations of CSCs and SFCs and the router switching capacity that results from these configurations.

Table 1-7 Switch Planes Provided by Switch Cards

Switch Card Type
Availability
Number of Switch Planes
Description

One CSC

Standard

1

A single CSC supports an OC-12 data rate for the router, but provides no redundancy in the router's switch fabric.

Second CSC

Optional

1

A second CSC supports an OC-12 data rate for the router and also provides a redundant plane of switch fabric. If one of the CSCs fails, a fault recovery cutover to the surviving CSC occurs, not only to maintain the router's OC-12 data rate, but also to preserve the system services peculiar to the CSC.

Three SFCs

Optional

3

The optional set of three SFCs enables the router to support an OC-48 data rate. In an OC-48 rate1 system, no redundancy exists in the switch fabric. However, if a switch plane failure occurs in a fully-redundant2 system, a CSC can take over the functions of either a failed CSC or a failed SFC, not only to maintain the router's OC-48 data rate, but also to preserve the essential CSC system services.

1 Router equipped with one CSC and three SFCs.

2 Router equipped with two CSCs and three SFCs.


Switch Fabric Controller Functions of the CSC

In addition to providing one plane of switch fabric for the router, the CSC provides numerous other functions and services essential to router operations. illustrates the primary functional elements of the CSC.

Figure 1-13 Block Diagram of the CSC

The major functions of each element of the CSC are summarized briefly in the following paragraphs.

Master clock generator—This function on the CSC provides a clock source to the RP, all installed switch cards (including the SFCs and a redundant CSC), and all installed line cards. The master clock generator synchronizes the transfer of data through the router's switch fabric.

In a redundant CSC configuration, the phase of the master clock generator on one card is synchronized with that of the other card. If either clock drifts, the master clock generators on both cards remain tightly aligned.

Should one of the CSCs fail, the phase lock between the two master clock sources is aborted within nanoseconds, enabling the surviving CSC clock to remain stable and take over master clock duties.

Frame synchronization generator—This function on the CSC provides a periodic signal to the line cards and switch cards to control data flow.

In a redundant CSC configuration, either CSC can adopt the frame synchronization phase of the other to ensure phase alignment. The line cards can switch between frame synchronization masters without disruption.

If the frame synchronization function on one CSC fails, cutover to the surviving frame synchronization generator on the other card occurs within nanoseconds, sustaining system operations.

Central switch allocator and scheduler—This function on the CSC allocates switching resources to line cards and schedules (arbitrates) the flow of data through the router's switch fabric.

Switch arbitration begins with a set of requests from line cards to send data through the router's switch fabric. The scheduler plans a set of paths through the switch fabric to carry as much data as possible per unit of time. At the next available time unit, the request to send data is granted, and the data is sent to its destination. The next round of switch arbitration (scheduling) then begins.

The scheduler also sends switch fabric control information to each switch plane to create appropriate data paths through the switch fabric. When the new data paths are configured into the switch fabric, data begins to flow toward the destination line card(s).

The central switch allocator and scheduler accepts data transport requests from all line cards (including the RP), generates grants (accepted data transport requests), and drives all planes of the router's switch fabric.

Single plane switch fabric—The CSC's single-plane switch fabric provides an OC-12 rate of switching capacity for the router. This switch fabric plane operates under control of the CSC's central switch allocator and scheduler.

This single switch plane of the CSC can be used alone in a minimum router configuration, or it can be used in combination with another CSC and the three optional SFCs for full switching redundancy. In the latter case, the per line-card slot bandwidth of the router is increased from an OC-12 rate to an OC-48 rate, and the second CSC provides redundancy.

If any one switch plane fails in a fully redundant switch fabric, the failed plane is shut down, and the router's full data bandwidth is carried by the surviving planes. The fault recovery cutover to another viable switch plane typically occurs without loss of data, because the data path defect is detected while redundancy information is still available, thus enabling error packets to be repaired "on the fly."

Housekeeping and Alarm Monitoring Functions of the CSC

The section describes the following housekeeping and alarm monitoring facilities built into the CSC:

MBus module— The MBus module on the CSC is a microprocessor-based subassembly that provides housekeeping services required during router power up and initialization. It also supports the alarm monitoring LEDs on the CSC faceplate, as described in the following section.

The MBus module on the CSC operates partly autonomously and partly under the control of the master MBus module on the RP.

A failed MBus module on the CSC is detected by an MBus polling algorithm running in the background on the RP.

A failed MBus module detected by this polling algorithm in a redundant CSC configuration causes the master MBus module to execute an administrative cutover to the MBus module on the surviving CSC. This cutover is accomplished with no disruption of normal system operations.

Alarm monitoring and status functions—The CSC supports the router's alarm and status monitoring functions. These functions are described in the following paragraphs. shows the location of the alarm contact connector and the various LEDs on the CSC faceplate through which the system accomplishes its alarm monitoring and status reporting functions.

Figure 1-14 CSC Alarm Monitoring Facilities

DB-25 alarm contact connector—A female DB-25 D-sub connector incorporated into the CSC faceplate enables you to attach an external alarm monitoring facility to the router, thus supporting a telco style of handling alarm conditions in the router.

The alarm signals sent to this DB-25 connector are identical in function to those sent to the system LEDs on the CSC faceplate (see the following section entitled "System Alarm LEDS").

Any alarm condition in the router that activates one of the system alarm LEDs on the CSC faceplate also energizes an appropriate CSC relay, causing a corresponding signal to be sent to the DB-25 connector. If an external alarm monitoring facility is attached to the DB-25 connector, this signal activates the appropriate external audible or visible alarm.

An external audible alarm can be reset by clearing the condition that caused the alarm or by pressing the alarm cutoff reset/lamp test (ACO/LT) button on the CSC faceplate (see the following discussion about the ACO/LT button). A visual alarm, however, can be reset only by resolving the problem that caused the alarm condition.

Only safety extra-low voltage (SELV) external alarm circuits can be connected to the DB-25 connector.

One external closure sense line, enabling the router to monitor an external event, such as the opening of a cabinet door or the activation of an alarm in associated equipment.

Alarm cutoff reset/lamp test (ACO/LT) button—If you equip your system with an external alarm monitoring facility, a visible indication can be provided, and/or an audible alarm can be sounded, to immediately notify site personnel of an alarm condition in the router.

An audible alarm generated by the system continues to sound until you either clear the alarm condition itself or press the ACO/LT button to silence the alarm. Merely pressing this button does not resolve the alarm condition.

You can test the operability of the LEDs on the CSC(s), the SFC(s), and the power supply(ies), by pressing the ACO/LT button at any time. Doing so causes the LEDs on all these router components to remain lit as long as you hold down the button. However, the LEDs on the SFCs are visible only when the air filter assembly is removed.

In a system equipped with two CSCs, pressing the ACO/LT button on one CSC is equivalent to pressing this button on either CSC or both CSCs.

System alarm LEDs—Three system LEDs, labeled critical, major, and minor, are incorporated into the CSC faceplate (see ) to signal the existence of alarm conditions detected in the router by the system's environmental monitoring circuitry.

During an alarm condition, one of these LEDs goes on to indicate the severity of the detected fault. During a critical alarm, the top LED (Critical) on the CSC faceplate indicates red; similarly, during a major alarm, the middle LED (Major) on the CSC faceplate also indicates red, signifying an alarm condition of lesser severity; finally, during a minor system alarm, the bottom LED (Minor) indicates amber, signifying an alarm condition of least severity.

At the same time that one of these LEDs goes on to signal the alarm event, an associated alarm relay on the CSC is closed, sending a corresponding signal to the DB-25 alarm contact connector on the CSC faceplate.

An alarm condition detected in a redundant CSC configuration causes the appropriate relays on both CSCs to close, activating the visible and audible alarm functions of the DB-25 connector on each card.

When the fault condition is resolved, MBus software running in the GRP automatically clears the fault indication by communicating with the master MBus module, which, in turn, communicates with the MBus module on each circuit board.

CSC Status LEDs—Two LEDs on the CSC faceplate, the top one labeled FAIL and the bottom one labeled ENABLED, indicate the operational status of the CSC.

FAN FAIL Status LEDs for each fan tray—Two side-by-side LEDs on the CSC faceplate indicate the operational status of the fan trays.

The LINECARD LED on the left pertains to the card cage fan tray, and the PWR SPLY LED on the right pertains to the power supply fan tray.

SFC Status LEDs—Two LEDs at the bottom of the CSC faceplate, the top one labeled FAIL and the bottom one labeled ENABLED, indicate the operational status of the SFCs in the lower card cage (behind the air filter assembly).

If the FAIL LED goes on, it indicates that one of the three SFCs in the lower card cage has failed. To determine which of the SFCs has failed, you must remove the air filter assembly and examine the status of the LEDs on each SFC.

Two side-by-side LEDS behind a vertical tab near the center of the SFC (see ) indicate the operational status of the card.

Figure 1-15 Status LEDs on an SFC

Board Power and Fan Tray Power Functions of the CSC

DC-DC converters on the CSC provide power for its own circuitry, as well as power for the fan trays. These functions are described briefly in the following sections.

Board power—A DC-DC converter on the CSC takes the -48 VDC being delivered to the card from the backplane and converts it into the +3.3 VDC required to drive the card's electronic circuity.

No redundancy is built into the CSC for the +3.3 VDC operating voltage; if the DC-DC converter fails to deliver this voltage, the card shuts down, at which time the redundant CSC, if installed, takes over to maintain normal system operations.

However, in a nonredundant CSC configuration, the failure of the installed CSC causes the entire system to shut down.

Fan tray power—The Cisco 12008 router contains two fan trays (see ).

Control of fan power is initiated at system startup, with the fans running at a slow rate for normal operations. Such operation minimizes fan noise, wear, and power consumption. A DC-DC converter on the CSC provides +20 VDC for slow fan operation and +25 VDC for fast fan operation when an overtemperature condition is sensed in the router.

Periodically, the master MBus module on the GRP polls the MBus module on each circuit board to determine whether router components are cool enough to warrant keeping the fans running at their minimum rate. If they are not, the master MBus module directs the MBus module on the CSC to increase the operating voltage being delivered to the fan trays, causing the fans to run faster, thus increasing the volume of air being circulated through the router.

Each fan is monitored separately for failure. A failed fan is not "shut off" in the usual sense; rather, a current-limiting feature in the faulty fan prevents it from interfering with the operation of other fans.

On failure of a fan in either the card cage fan tray or the power supply fan tray, the CSC increases the voltage being delivered to the surviving fans, causing them to run faster to compensate for the failed fan.

Cisco 12000 Series Line Cards

The Cisco 12008 comes equipped with the number and type of line cards that you ordered already installed. Up to seven Cisco 12000 series line cards can be installed in the router to support a variety of physical network media.

The line cards can be installed in upper card cage slots 0 through 3 and slots 4 through 7. Note, however, that it is recommended that the GRP be installed in slot 0. Line cards interface to each other and the GRP through the router's switch fabric.

The following types of line cards are available for use with the Cisco 12008:

Quad OC-3c/STM-1c POS—4 ports

OC-12c/STM-4c POS—1 port

OC-12c/STM-4c ATM—1 port

These cards provide the interfaces to the router's external physical media. They exchange packet data with each other by way of the router's switch fabric.


Caution   
Any unoccupied slot in the upper card cage must have a blank filler panel installed for EMI compliance and to ensure proper air flow through the router enclosure.

A vertical cable-management bracket attached to the faceplate of each line card enables you to neatly arrange the network interface cables for connection to the individual ports on the line card. The cable-management system is described in detail in the section entitled "Cable-Management System" on page 8.

The online insertion and removal (OIR) capability of the Cisco 12008 enables you to remove and replace a line card while the system remains powered up and operational.

The Cisco 12000 series line cards available for use with the Cisco 12008 router are described briefly in the following sections.

Quad OC-3c/STM-1c POS Line Card

The Quad OC-3c/STM-1c POS line card provides the Cisco 12008 router with four independent Packet-Over-SONET (POS) ports on a single card. The card interfaces with the router's switch fabric and provides four OC-3c/STM-1c SC-duplex SONET connections. These connections are concatenated, which provides for increased efficiency by eliminating the need to partition the bandwidth.

shows a high-level block diagram of the Quad OC-3c/STM-1c POS line card; shows a front view of the card.

Figure 1-16 Block Diagram of the Quad OC-3c/STM-1c POS Line Card

Figure 1-17 Quad OC-3c/STM-1c POS Line Card

Each Quad OC-3c/STM-1c POS line card incorporates the following major components:

Transceivers—The single-mode intermediate reach transceiver provides a full-duplex, 155-Mbps, 1300-nm, laser-based SONET/SDH-compliant interface. The multimode transceiver provides a full-duplex, 155-Mbps, 1300-nm, LED-based SONET/SDH  compliant interface.

The SONET specification for fiber-optic transmission defines two types of fiber: single mode and multimode. Signals can travel farther through single mode fiber than through multimode fiber.

The maximum distance for single-mode installations is determined by the amount of light loss in the fiber path. Good quality single-mode fiber with very few splices can carry an OC-3c/STM-1c signal 9.3 miles (15 km) or more; good quality miltimode fiber can carry a signal up to 1.3 miles (2 km).

Burst buffers—The Quad-OC3c/STM-1c contains four 128-KB burst buffers. The burst buffer prevents the dropping of packets during instantaneous increases in the number of back-to-back small packets being transmitted at OC-3 line rates.

Burst buffers are used to achieve high throughput while smoothing out the arriving packet burst for the Layer 3 switch processor.

Buffer memory—The silicon queuing engine controls the placement of IP packets in buffer memory as well as their removal from buffer memory. The default packet buffer memory is 32 MB, which includes 16 MB of receive (Rx) buffers and 16 MB of transmit (Tx) buffers.

The buffer memory can be configured to support up to 64 MB of receive buffers and up to 64-MB of transmit buffers. The buffers can support delays comparable to the longest round trip delays measured in the Internet at OC-3c/STM-1c line rates.

Layer 2 switching accelerator—The Layer 2 switching accelerator assists the forwarding processor. It is a specially designed application-specific integrated circuit (ASIC) that optimizes access to the Layer 2 and Layer 3 information within each packet. At very high line rates, this access process must be executed as rapidly as possible, which is why an ASIC is dedicated to the process.

Forwarding processor—A forwarding processor makes forwarding decisions based on the information in the Cisco Express Forwarding (CEF) table and the Layer 2 and Layer 3 information in the packet. The GRP constantly updates forwarding information in the forwarding table based on the latest information in the routing table.

Once the forwarding decision has been made, the silicon queuing engine is notified by the forwarding processor, and the silicon queuing engine places the packet in the proper queue.

This partitioning between the Layer 2 switching accelerator and the forwarding processor blends the high throughput of hardware-accelerated forwarding with the flexibility of software-based routing.

Silicon queuing engine—Each line card has two silicon queuing engines: receive and transmit. The receive engine moves packets from the burst buffer to the switch fabric, and the transmit engine moves packets from the switch fabric to the transmit interface.

When an incoming IP packet is clocked into the silicon queuing engine, the packet's integrity is verified by a check of the CRC. Next, the silicon queuing engine transfers the IP packet to buffer memory and tells the Layer 3 switching accelerator the location of the IP packet.

Simultaneously, the silicon queuing engine is receiving forwarding information from the forwarding processor. The forwarding processor tells the silicon queuing engine the virtual output queue where the IP packet is to be placed.

Each virtual output queue represents an output destination (destination line card). This placement of the IP packets in a virtual output queue is based on the decision made by the forwarding processor. There is one virtual output queue for each line card, plus a dedicated virtual output queue for multicast service.

The transmit silicon queuing engine moves the packet from the switch fabric to the transmit buffer, and then to the transmit interface.

Switch fabric interface—The switch fabric interface is the same 1.25-Gbps, full-duplex data path to the switching fabric that is used by the GRP. Once a packet is in the proper queue, the switch fabric interface issues a request to the master clock scheduler on the CSC. The scheduler issues a grant and transfers the packet across the switching fabric.

Maintenance bus (MBus) module—A maintenance bus (MBus) module on the line card responds to requests from the master MBus module on the GRP. The MBus module on the line card reports temperature and voltage information to the master MBus module. In addition, the MBus module on the line card contains the ID-EEPROM, which stores the serial number, hardware revision level, and other information about the card.

Cisco Express Forwarding (CEF) memory table—Each line card maintains CEF tables. These tables, derived from routing tables maintained by the GRP, are used by the line card processor in making forwarding decisions.

Large networks may require more DRAM to support large CEF tables. For information on adding memory to a line card, see the document entitled Cisco 12000 Series Gigabit Switch Router Memory Replacement Instructions.

OC-12c/STM-4c POS Line Card

The OC-12c/STM-4c POS line card provides the Cisco 12008 with a single 622-Mbps Packet-Over-SONET (POS) interface. The card provides one OC-12c/STM-4cc SC duplex single-mode or multimode SONET/SDH connection. This connection is concatenated, which provides for increased efficiency by eliminating the need to partition the bandwidth.

shows a high-level block diagram of the OC-12c/STM-4c POS line card; shows a front view of the card.

Figure 1-18 Block Diagram of the OC-12c/STM-4c POS Line Card

Figure 1-19 OC-12c/STM-4c POS Line Card

Each OC-12c/STM-4c POS line card incorporates the following primary components:

Each Quad OC-3c/STM-1c POS line card incorporates the following major components:

Transceivers—The single-mode intermediate reach transceiver provides a full-duplex, 155-Mbps, 1300-nm, laser-based SONET/SDH-compliant interface. The multimode transceiver provides a full-duplex, 155-Mbps, 1300-nm, LED-based SONET/SDH  compliant interface.

The SONET specification for fiber-optic transmission defines two types of fiber: single mode and multimode. Signals can travel farther through single mode fiber than through multimode fiber.

The maximum distance for single-mode installations is determined by the amount of light loss in the fiber path. Good quality single-mode fiber with very few splices can carry an OC-3c/STM-1c signal 9.3 miles (15 km) or more; good quality miltimode fiber can carry a signal up to 1640 feet (500 m).

Burst buffers—The burst buffer (512 KB) prevents the dropping of packets during instantaneous increases in the number of back-to-back small packets being transmitted at OC-12c/STM-4c line rates. Burst buffers are used to achieve high throughput while smoothing out the arriving packet burst for the Layer 3 switch processor.

Buffer memory—The silicon queuing engine controls the placement of IP packets in buffer memory as well as their removal from buffer memory. The default packet buffer memory is 32 MB, which includes 16 MB of receive (Rx) buffers and 16 MB of transmit (Tx) buffers.

The buffer memory can be configured to support up to 64 MB of receive buffers and up to 64 MB of transmit buffers. The buffers can support delays comparable to the longest round trip delays measured in the Internet at OC-12c/STM-4c line rates

Layer 2 switching accelerator—The Layer 2 switching accelerator assists the forwarding processor. It is a specially designed application-specific integrated circuit (ASIC) that optimizes access to the Layer 2 and Layer 3 information within each packet. At very high line rates, this access process must be executed as rapidly as possible, which is why an ASIC is dedicated to the process.

Forwarding processor—A forwarding processor makes forwarding decisions based on the information in the Cisco Express Forwarding (CEF) table and the Layer 2 and Layer 3 information in the packet. The GRP constantly updates forwarding information in the forwarding table, based on the latest information in the routing table.

Once the forwarding processor makes a forwarding decision, it notifies the silicon queuing engine, and the silicon queuing engine places the packet in the proper queue.

This partitioning between the Layer 2 switching accelerator and the forwarding processor blends the high throughput of hardware-accelerated forwarding with the flexibility of software-based routing.

Silicon queuing engine—Each line card has two silicon queuing engines: receive and transmit. The receive engine moves packets from the burst buffer to the switch fabric, and the transmit engine moves packets from the switch fabric to the transmit interface.

When an incoming IP packet is clocked into the silicon queuing engine, packet integrity is verified by a CRC check. Next, the silicon queuing engine transfers the IP packet to buffer memory and tells the Layer 3 switching accelerator the location of the IP packet. Simultaneously, the silicon queuing engine is receiving forwarding information from the forwarding processor. The forwarding processor tells the silicon queuing engine the virtual output queue where the IP packet is to be placed.

Each virtual output queue represents an output destination (destination line card). This placement of the IP packets in a virtual output queue is based on the decision made by the forwarding processor. There is one virtual output queue for each line card, plus a dedicated virtual output queue for multicast service.

The transmit silicon queuing engine moves the packet from the switch fabric to the transmit buffer, and then to the transmit interface.

Switch fabric interface—The switch fabric interface is the same 1.25-Gbaud, full-duplex data path to the switching fabric that is used by the GRP. Once a packet is in the proper queue, the switch fabric interface issues a request to the master clock scheduler on the CSC. The scheduler issues a grant and transfers the packet across the switching fabric.

Maintenance bus (MBus) module—A maintenance bus (MBus) module on the line card responds to requests from the master MBus module on the GRP. The MBus module on the line card reports temperature and voltage information to the GRP master MBus module.

In addition, the MBus module on the line card contains the ID-EEPROM, which stores the serial number, hardware revision level, and other information about the card.

Cisco Express Forwarding (CEF) memory table—Each line card maintains CEF tables. These tables, derived from routing tables maintained by the GRP, are used by the line card processor to make forwarding decisions.

Large networks may require more DRAM to support large CEF tables. For information on adding memory to a line card, see the document entitled Cisco 12000 Series Gigabit Switch Router Memory Replacement Instructions.

OC-12c/STM-4c ATM Line Card

The OC-12c/STM-4c ATM line card provides the Cisco 12008 with a 622-Mbps ATM interface. The card interfaces to the router's switch fabric, supports from 10 to 40 Gbps, and provides one OC-12c/STM-4c SC duplex single-mode or multimode SONET/SDH connection. This connection is concatenated, which provides for increased efficiency by eliminating the need to partition the bandwidth.

shows a high-level block diagram of the OC-12c/STM-4c ATM line card; shows a front view of the card.

Figure 1-20 Block Diagram of the OC-12c/STM-4c ATM Line Card

Figure 1-21 Front View of OC-12c/STM-4c ATM Line Card

Each OC-12c/STM-4c ATM line card incorporates the following primary components:

Reassembly and segmentation—The transceivers support packet reassembly (converting ATM cells to packets) and segmentation (converting packets to ATM cells). The transceivers can handle up to 4000 simultaneous reassemblies (based on an average packet size of 280 bytes). In addition, the reassembly application-specific integrated circuit (ASIC) and the segmentation ASIC support up to 15,000 active virtual circuits.

The SONET specification for fiber-optic transmission defines two types of fiber: single mode and multimode. Signals can travel farther through single mode fiber than through multimode fiber.

The maximum distance for single-mode installations is determined by the amount of light loss in the fiber path. Good quality single-mode fiber with very few splices can carry an OC-3c/STM-1c signal 9.3 miles (15 km) or more; good quality miltimode fiber can carry a signal up to 1640 feet (500 m).

Burst buffers—The burst buffer (4 MB) prevents the dropping of packets during instantaneous increases in the number of back-to-back small packets being transmitted at OC-12 line rates. Burst buffers provide high throughput while smoothing out the arriving packet burst for the Layer 3 switch processor.

Buffer memory—The silicon queuing engine controls the placement of IP packets in buffer memory as well as their removal from buffer memory. The default packet buffer memory is 32 MB, which includes 16 MB of receive (Rx) buffers and 16 MB of transmit (Tx) buffers. The buffer memory can be configured to support up to 64 MB of receive buffers and 64 MB of transmit buffers. The buffers can support delays comparable to the longest round trip delays measured in the Internet at OC-3/STM-1 line rates.

Layer 2 switching accelerator—The Layer 2 switching accelerator assists the forwarding processor. It is a specially designed application-specific integrated circuit (ASIC) that optimizes access to the Layer 2 and Layer 3 information within each packet. At very high line rates, this access process must be executed as rapidly as possible, which is why an ASIC is dedicated to the process.

Forwarding processor—A forwarding processor makes forwarding decisions based on information in the Cisco Express Forwarding (CEF) table and the Layer 2 and Layer 3 information in the packet. The GRP constantly updates forwarding information in the forwarding table based on the latest information in the routing table.

Once the forwarding decision has been made, the silicon queuing engine is notified by the forwarding processor, and the silicon queuing engine places the packet in the proper queue.

This partitioning between the Layer 2 switching accelerator and the forwarding processor blends the high throughput of hardware-accelerated forwarding with the flexibility of software-based routing.

Silicon queuing engine—Each line card has two silicon queuing engines: receive and transmit. The receive engine moves packets from the burst buffer to the switch fabric, and the transmit engine moves packets from the switch fabric to the transmit interface.

When an incoming IP packet is clocked into the silicon queuing engine, the packet's integrity is verified by a CRC check. Next, the silicon queuing engine transfers the IP packet to buffer memory and tells the Layer 3 switching accelerator the location of the IP packet. Simultaneously, the silicon queuing engine is receiving forwarding information from the forwarding processor, while the forwarding processor is telling the silicon queuing engine where the IP packet is to be placed in the virtual output queue.

Each virtual output queue represents an output destination (destination line card). Placement of the IP packets in a virtual output queue is based on the decision made by the forwarding processor. There is one virtual output queue for each line card, plus a dedicated virtual output queue for multicast service.

The transmit silicon queuing engine moves the packet from the switch fabric to the transmit buffer, and then to the transmit interface.

Switch fabric interface—The switch fabric interface is the same 1.25-Gbps, full-duplex data path to the switching fabric that is used by the GRP. Once a packet is in the proper queue, the switch fabric interface issues a request to the master clock scheduler on the CSC. The scheduler issues a grant and transfers the packet across the switching fabric.

Maintenance bus (MBus) module—An MBus module on the line card responds to requests from the master MBus module on the GRP. The line card MBus module reports temperature and voltage information to the master MBus module.

In addition, the MBus module on the line card contains the ID-EEPROM, which stores the serial number, hardware revision level, and other information about the card.

Cisco Express Forwarding (CEF) memory table—Each line card maintains CEF tables. These tables, derived from routing tables maintained by the GRP, are used by the line card processor to make forwarding decisions.

Large networks may require more DRAM to support large CEF tables. For information on adding memory to a line card, see the document entitled Cisco 12000 Series Gigabit Switch Router Memory Replacement Instructions.

Air Filter Assembly

The Cisco 12008 is equipped with a removable air filter that is mounted directly to the router enclosure in front of the lower card cage (see ).

Although the Cisco 12008 will run without an air filter, the air filter should always be present and maintained properly, especially in dirty or dusty environments.

The air filter assembly serves the following purposes:

Filters the ambient air being draw into the router by the card cage fan tray.

Prevents EMI radiation from being emitted into the router's environment.

A metal honeycomb structure built into the air filter assembly provides EMI containment.

You are advised to inspect and clean the air filter at least once a month (or more often in a dusty environment).

Procedures for vacuuming and replacing the air filter are contained in the section entitled "Cleaning the Air Filter" in Chapter 7.

Lower Card Cage and Associated Components

The lower card cage, located directly behind the air filter assembly (see ), houses the card cage fan tray and an optional set of three switch fabric cards (SFCs).

The dimensional characteristics of the SFCs differ markedly from those of the circuit boards in the upper card cage. Three dedicated slots, numbered SFC0, SFC1, and SFC2 as you face the lower card cage, are provided to house the SFCs.

Switch Fabric Cards

The SFCs increase the switching capacity of the Cisco 12008. By adding three SFCs to a router equipped with a single CSC, you increase the bandwidth of each line card slot in the router from an OC-12 rate to an OC-48 rate.

By adding three SFCs to a router equipped with two CSCs, you not only increase the bandwidth of each line card slot to an OC-48 rate, but you also provide a fifth (redundant) switch plane so that the router's OC-48 data rate can be maintained even if a switch plane should fail.

In a router with full switch plane redundancy (that is, a router with five available switch planes), five parallel 1.25 Gbaud serial data streams can be transmitted across the backplane to and from the router's line cards. However, only four of the data streams are required for data transmission purposes; the fifth data stream carries error correction information. If an error occurs on one of the parallel data streams, data in error can be recovered through use of the four remaining correct data streams.

You need not install the optional SFCs in a router that uses line cards having an aggregate bandwidth rate of OC-12 or less. In such a system, a single CSC can provide sufficient bandwidth to accomplish all the router's switching and routing functions. Thus, a minimally configured router does not require the optional switching capacity provided by the SFCs. To increase the switching capacity of the Cisco 12008 to the full OC-48 rate, however, you must install the three optional SFCs.

Each SFC is mounted on its own card carrier and incorporates an onboard power supply that takes the -48 VDC supplied by the backplane and converts it into the 3.3 VDC operating voltage required by the card.

Figure 1-22 Components in the Lower Card Cage

The switching fabric of the SFC is identical to that of the CSC. However, the SFCs do not perform any of the system services native to the CSC (see the section entitled "Clock and Scheduler Card" on page 44). The SFC merely augments the switching capacity of the router.

Power Distribution System in the Cisco 12008

In the Cisco 12008, source AC or source DC power is converted by the installed power supply(ies) into the +5 VDC and -48 VDC required for router operation. These voltages are delivered to the backplane through the blind mating Elcon connector at the rear of the power supply enclosure. The backplane then distributes these operating voltages to all of the installed components in the system (see ).

The +5 VDC is fed to the MBus module on each installed card, and the -48 VDC is fed to a DC-DC converter on each card.

The DC-DC converter on each card operates under control of the card's MBus module. When directed by the GRP or system software during normal system startup, the DC-DC converter on each card is activated to convert the -48 VDC from the backplane into the voltages required to power the card's electronic circuitry.

The card cage fan tray and the power supply fan tray derive their operating power from a DC-DC converter on the CSC. This converter takes the -48 VDC from the backplane and converts it into the +24 VDC operating voltage required by the fan trays.

If an overtemperature condition is sensed anywhere within the router, or if any one of the fans fails in either the card cage fan tray or the power supply fan tray, the DC-DC converter on the CSC increases the voltage being delivered to the fan trays. This causes the fans to run at maximum speed to increase the volume of cooling air flowing through the router. Once the overtemperature condition is resolved, the fans revert to their normal operating speed.

Because the fans must operate continuously to prevent thermal damage to router components, they cannot be turned off by software.

Figure 1-23 Power Distribution System in the Cisco 12008

Cisco 12008 Environmental Monitoring Facility

An environmental monitoring facility, called the maintenance bus (MBus), supports a variety of functions essential to router operations. These functions include the following:

System discovery (enabling the router to identify installed components)

Booting software images

Supporting console traffic, logging functions, and diagnostic functions

Monitoring the operational health of the router and reporting error conditions

The MBus facility in the router is interconnected by means of the backplane to the following components:

GRP

Line cards

CSCs

SFCs

Power supplies

Each of the components listed here contains an onboard MBus module that incorporates two separate transceivers (A and B). Each transceiver has a separate etch (communication path) through the backplane. Consequently, all the MBus modules in the system are reliably interconnected to each other by means of redundant busses. This redundancy enhances the reliablity of the entire environmental monitoring system.

The MBus module on each component is powered by +5.2 VDC that it receives through the backplane from the power supply. A single MBus firmware image executes on all the MBus modules present in the system.

The master MBus module on the GRP monitors all the alarm conditions detected by the MBus modules in the other components of the system. The master MBus module then determines an appropriate response to the alarm condition.

The MBus modules on installed components perform the following functions:

Power-up/down control—When power is applied to the router, the MBus module on the GRP and the CSC immediately receive +5.2 VDC through the backplane from the power supply, causing each card to supply power to its circuitry.

The MBus modules on other installed components then power up on command from the master MBus agent on the GRP.

Device discovery—The GRP determines the system configuration by means of the MBus facility.

A message is sent from the master GRP MBus agent, requesting that all installed components identify themselves. Each return response includes slot number, card type, and component type.

Downloading software—A line card ROM monitor is loaded into Flash ROM on the card during the manufacturing process. This image, which can be field upgraded, if necessary, boots software to the line card by means of the MBus facility.

Because the MBus is slow relative to the switch fabric, only enough code is initially downloaded to the line card to enable it to access the router's switch fabric.

This initial code includes a line card fabric downloader that functions as a secondary bootstrap program to quickly complete the downloading of the Cisco IOS image to the line card by means of the router's high-speed switch fabric.

Diagnostics—The MBus facility enables field diagnostics to be run on the GRP and the line cards, whether the router is in service (running diagnostics on an individual card without taking the router offline) or out of service (taking the entire router down to run diagnostics).

Environmental monitoring and alarm functions—The environmental monitoring functions of the MBus system include the following:

Voltage and temperature monitoring for the router's installed components

Fan failure sensing for the card cage fan tray and the power supply fan tray

System Specifications

lists the physical specifications of the Cisco 12008.

outlines the electrical specifications of the AC-input power supply; outlines similar specifications for the DC-input power supply.

lists the environmental specifications of the Cisco 12008.

Table 1-8 Physical Specifications of the Cisco 12008

Description
Value

Chassis height

24.8 inches (63.6 cm)

Chassis width

17.4 inches (44.6 cm)

19.1 inches (48.5 cm), including mounting flanges

Chassis depth

21.2 inches (54.4 cm), including cable- management system

Weight, maximum configuration

180 lb (81.7 kg) with two DC-input power supplies

187 lb (84.9 kg) with two AC-input power supplies

Weight, minimum configuration

127 lb (57.7 kg)

Weight, shipping pallet

44 lb (20 kg)

Weight, total system, on pallet

231 lb (104.9 kg)

Weight, base chassis with backplane

50 lb (22.7 kg)

Weight, card cage fan tray

12 lb (5.4 kg)

Weight, power supply fan tray

2 lb (0.9 kg)

Weight, AC-input power supply

17 lb (7.7 kg)

Weight, DC-input power supply

14 lb (6.4 kg)

Weight, line card

8 lb (3.6 kg)

Weight, GRP

8 lb (3.6 kg)

Weight, CSC

7 lb (3.2 kg)

Weight, SFC

2 lb (0.9 kg)


Table 1-9 Electrical Specifications of the AC-Input Power Supply

Power Supply
Type
Electrical Characteristic
Value

AC

Input power

Maximum: 2000W
200 VAC to 240 VAC @ 10A

AC

Input voltage

Nominal: 200 VAC to 240 VAC, single phase
Tolerance limits: 180 VAC to 264 VAC

AC

Input current

9.5A @ 200 VAC

AC

Line frequency

47 to 63 Hz

AC

Output power

Maximum: 1560W
-48 VDC @ 33.7A
+5 VDC @ 20.8A)


Table 1-10 Electrical Specifications of the DC-Input Power Supply

Power Supply
Type
Electrical Characteristic
Value

DC

Input power

Maximum: 1580W
-40.5 VDC to -75 VDC @ 39A to 21A

DC

Input voltage

Nominal: -48 VDC (United States)
Tolerance limits: -40.5 VDC to -56 VDC
Nominal: -60 VDC (International)
Tolerance limits: -58 VDC to -75 VDC

DC

Input current

33.75A maximum @ -48 VDC
27A maximum @ -60 VDC
Internal circuit breaker is rated at 40A

DC

Output power

Maximum: 1542W
-48 VDC @ 33.7A
+5 VDC @ 20.8A


Table 1-11 Environmental Specifications of the Cisco 12008

Description
Value

Temperature

Operating: 32° to 104° F (0° to 40° C)
Nonoperating: -4° to 149° F (-20° C to 65° C)

Humidity

Noncondensing, operating: 10 to 90%
Noncondensing, nonoperating: 5 to 95%

Altitude

Operating: 0 to 10,000 ft (0 to 3048 m)
Nonoperating: 0 to 30,000 ft (0 to 9144 m)

Heat dissipation

6,000 Btu/hr maximum

Acoustic Noise

69 dbA maximum

Shock

Operating: 5 to 500 Hz, 0.5 g1 (0.1 oct/min2 )
Nonoperating: 5 to 100 Hz, 1 g (0.1 oct/min);
100 to 500 Hz, 1.5g (0.2 oct/min);
500 to 1000 Hz, 1.5 g (0.2 oct/min)

1 g = gravity.

2 oct/min = octave per minute.


Agency Approvals

In addition to meeting GR-63-CORE and GR-1089-CORE specifications, the Cisco 12008 meets the requirements of the agencies listed in .

Table 1-12 Agency Approvals

Category
Agency Approval

Safety

UL 1950

 

CSA 22.2 No. 950

 

EN60950

 

AUSTEL TS001

 

AS/NZS 3260

EMI

FCC Class A

 

CSA Class A

 

EN55022 Class A

 

VCCI Class 2

 

AS/NRZ 3548 Class A

Immunity

EN61000-4-2/IEC-1000-4-2

 

EN61000-4-3/IEC-1000-4-3

 

EN61000-4-4/IEC-1000-4-4

 

EN61000-4-5/IEC-1000-4-5

 

EN61000-4-6/IEC-1000-4-6

 

EN61000-4-11/IEC-1000-4-11