Cisco CMTS Feature Guide - N+1 Redundancy for the Cisco Cable Modem Termination System

Table Of Contents

N+1 Redundancy for the Cisco Cable Modem Termination System

Contents

Prerequisites

Restrictions and Limitations

General N+1 Redundancy Restrictions and Limitations

N+1 Redundancy Restrictions and Requirements for the Cisco uBR7246VXR Router

N+1 Redundancy Restrictions and Requirements for the Cisco uBR10012 Router

Information About N+1 Redundancy and the Cisco Universal Broadband CMTS

The Components and Terminology of N+1 Redundancy

N+1 Redundancy on the Cisco uBR10012 Universal Broadband Router

N+1 Redundancy on the Cisco uBR7246VXR Universal Broadband Router

N+1 Redundancy and the Cisco RF Switches

IF Muting on the Cisco CMTS for non-SNMP-capable Upconverters

Restrictions for IF Muting

Requirements for IF Muting

DSX Messages and Synchronized PHS Information

High Availability Support for Encrypted IP Multicast

Manual RF Switch Configuration Tasks for N+1 Redundancy

Configuring the Cisco RF Switch for N+1 Redundancy

Creating Cisco RF Switch Module Bitmaps

Global N+1 Line Card Redundancy

Configuring the Cisco uBR10012 Universal Broadband Router for Global N+1 Line Card Redundancy

Default Line Card and Bitmap Settings on the Cisco RF Switch for Global 7+1 Line Card Redundancy

Changing Default RF Switch Subslots for N+1 Line Card Redundancy

Displaying Global N+1 Line Card Redundancy Configuration

Configuring DHCP on the Cisco uBR10012 Universal Broadband Router to Assign IP Addresses on the Cisco RF Switch

Using Optional RF Switch Settings with Global N+1 Redundancy

Using Line Card Switchover and Revertback Commands for Global N+1 Redundancy

Using HCCP Lock and Unlock for Global N+1 Redundancy

How to Configure N+1 Redundancy on the Cisco CMTS

Preconfiguring HCCP Protect Interfaces for N+1 Redundancy

Operating DHCP with the Cisco RF Switch

Configuring HCCP Groups for Legacy N+1 Line Card Redundancy

Enabling HCCP Protect Interfaces for N+1 Redundancy

Configuring Global HCCP 4+1 and 7+1 Line Card Redundancy on the Cisco uBR10012 Router

Prerequisites

Restrictions

Examples

What to Do Next

Enabling the HCCP Switchover Enhancements Feature

Virtual Interface Bundling

Prerequisites for Enabling the HCCP Switchover Enhancements Feature

Maintaining Online Cable Modem Service When Removing HCCP Configuration from Working HCCP Interfaces

Shutting Down HCCP Protect Interfaces

Locking out HCCP Interface Switchover

Removing HCCP Configuration from HCCP Working or HCCP Protect Interfaces

Switchover Testing Tasks for N+1 Redundancy

Pre-testing System Check Procedures

Displaying HCCP Group Status on the Cisco CMTS

Displaying HCCP Working and HCCP Protect Interface Status

Displaying Cisco RF Switch Module Status on the Cisco RF Switch

Switchover Testing Procedures

Testing Cisco RF Switch Relays with Manual Switchover

Testing HCCP Groups with Manual Switchover

Using the show cable modem Command After a Manual Switchover

Background Path Testing for HCCP N+1 Redundancy on the Cisco uBR10012 Universal Broadband Router

Configuration Examples for Cisco N+1 Redundancy

Example: Cisco 3x10 RF Switch Modules in 8+1 Mode

Example: Cisco 3x10 RF Switch Modules in 4+1 Mode

N+1 Configuration Example on the Working Cisco uBR7246VXR Router

N+1 Configuration Example on the Protect Cisco uBR7246VXR Router

Examples: Cisco 3x10 RF Switch with Cisco uBR10012 Chassis

HCCP Working 1 Example

HCCP Working 2 Example

HCCP Working 3 Example

HCCP Working 4 Example

HCCP Protect Interface Configuration Examples

Example: Channel Switch Information from the Cisco uBR10012 Router

Example: Cisco 3x10 RF Switch and Cisco uBR10012 Chassis

Example: Cisco 3x10 RF Switches and Cisco uBR10012 Chassis

Example: Cisco 3x10 RF Switches and uBR7246VXR Chassis

HCCP Working uBR7246VXR Chassis 1

HCCP Protect uBR7246VXR Chassis

Additional References

Related Documents

Standards

MIBs

RFCs

Technical Assistance

N+1 Redundancy for the Cisco Cable Modem Termination System

Revised: June 6, 2007, OL-1467-08

This chapter provides procedures and commands by which to configure the N+1 Redundancy feature on the Cisco Cable Modem Termination System (CMTS), using the Cisco uBR10012 universal broadband routers with the Cisco 3x10 RF Switch.

N+1 redundancy refers to (N) cable interface line cards, called "Working" line cards being protected by one additional line card (+1), called the "Protect" line card. N+1 redundancy, of which 4+1 redundancy is one version, is made possible with the addition of a single Cisco RF Switch to your cable headend network. Together with the Cisco uBR10012 router, the Cisco RF Switch provides a fully redundant system that enables cable operators to achieve PacketCable system availability, minimize service disruptions, and simplify operations.

N+1 redundancy is an important step toward high availability on CMTS and telecommunications networks that use broadband media. N+1 redundancy can help limit Customer Premises Equipment (CPE) downtime by enabling robust automatic switchover and recovery in the event that there is a localized system failure.

Beginning with Cisco IOS Release 12.2(15)BC2a, N+1 redundancy adds synchronization between Hot-Standby Connection-to-Connection Protocol (HCCP) working interface configurations and those inherited upon switchover to HCCP protect interfaces. This makes the configuration of both easier and switchover times faster.

Global N+1 Line Card Redundancy, or HCCP Rapid Configuration, is a feature that simplifies the configuration of Working and Protect interfaces by eliminating the need to configure the more complex hccp interface configuration commands. Global N+1 Line Card Redundancy is supported on the Cisco uBR10012 router only with the Cisco UBR-MC5X20S, Cisco UBR10-MC5X20U, and Cisco UBR10-MC5X20H broadband processing engines (BPEs). Support for global 7+1 redundancy was introduced in Cisco IOS Release 12.3(13a)BC. In Cisco IOS Release 12.3(17a)BC, global N+1 redundancy was extended to support 4+1 configurations.

Beginning in Cisco IOS Release 12.3(21)BC, the Cisco uBR10012 universal broadband router supports the HCCP Switchover Enhancements feature that implements performance improvements for traffic recovery during line card switchover under certain scalability limits. For networks with less than 5000 cable modems per line card, and less than 1000 voice calls per line card, these switchover improvements include under 1-second recovery for voice calls, and under 20-second recovery for data traffic. In addition, the keepalive failure logic is modified to improve false switchovers.

Cisco IOS and Cisco RF Switch Firmware for N+1 Redundancy

Two operating systems govern the configuration and operation of N+1 Redundancy on the Cisco CMTS:

•Cisco Internetwork Operating System (IOS)—Governs the configuration and operation of Cisco universal broadband routers, and works closely with Cisco RF Switch Firmware when configured in N+1 Redundancy.

Note The Cisco IOS CLI now synchronizes configurations between HCCP Working and Protect interfaces. Preconfiguration of the Protect interfaces is no longer required in most circumstances.

•Cisco RF Switch Firmware—Governs the configuration and operation of the Cisco RF Switch, including the IP address on the RF Switch.

Both command-line interfaces above are required for configuration and testing of N+1 Redundancy.

Cisco IOS Feature Specifications for N+1 Redundancy on the Cisco Cable Modem Termination System

Release

Modification

12.1(10)EC

HCCP support introduced on the Cisco uBR7200 series routers.

12.2(4)XF1, 12.2(4)BC1

HCCP N+1 Redundancy support was added for the Cisco uBR10012 router and UBR10-LCP2-MC28C cable interface line card.

12.2(8)BC2

HCCP N+1 Redundancy support was added for the Cisco uBR10012 router and Cisco uBR10-LCP2-MC16x cable interface line cards.

12.2(11)BC1

HCCP N+1 Redundancy support was added for the Cisco uBR7246VXR router and Cisco uBR-LCP-MC16x cable interface line cards.

12.2(15)BC1

HCCP N+1 Redundancy support introduced for the Cisco uBR10012 router and Cisco UBR10-MC 5X20U or -S broadband processing engine (BPE).

12.2(15)BC2a

•HCCP N+1 Redundancy support introduced for the Cisco uBR7246VXR router and the Cisco uBR 3x10 RF Switch.

•CLI Usability—Synchronizes HCCP interface command-line interface (CLI) configuration between Working and Protect interfaces.

•Support for N+1 Redundancy for the Cisco UBR10-MC 5X20U or -S BPE on the Cisco uBR10012 router.

•IF Muting on the Cisco CMTS for non-SNMP-capable Upconverters — enables N+1 Redundancy on CMTS headends that do not use SNMP-enabled upconverters.

12.3(13a)BC

HCCP N+1 Redundancy on the Cisco 7200 series routers is no longer supported.

The following enhancements were introduced to HCCP N+1 redundancy support on the Cisco uBR10012 router:

•Global N+1 Line Card Redundancy

•Automatic running of the show hccp channel switch command for Background Path Testing for HCCP N+1 Redundancy on the Cisco uBR10012 Universal Broadband Router

12.3(17a)BC

The following High Availability enhancements were introduced for the Cisco CMTS:

•Enhanced globally-configured N+1 Redundancy on the Cisco uBR10012 router:

–Added global 4+1 redundancy support to the existing global 7+1 redundancy on the Cisco uBR10012 router.

–Supporting redundancy and show command enhancements

•Encrypted IP Multicast is supported during High Availability switchover events.

•PHS rules synchronize and are supported during High Availability switchover events.

12.3(21)BC

The following support has been removed:

•HCCP N+1 Redundancy support is removed for the Cisco uBR7246VXR router.

•Tracking of HCCP interfaces is removed. The hccp track command is obsolete.

The HCCP Switchover Enhancements feature is introduced on the Cisco uBR10012 router, with the following new support:

•Performance improvements for traffic recovery during line card switchover under certain scalability limits. Within the required network scalability limits, the HCCP Switchover Enhancements feature provides the following switchover benefits:

–Less than 1-second voice call recovery.

–Less than 20-second data recovery.

•To prevent false switchovers, the keepalive failure logic is modified.

•For faster line card switchovers, the member subslot protect command has been modified to add the [config slot/subslot] option. When using the new config option, you can preload upstream connectors on an HCCP protected interface to emulate the most common line card connector assignments.

Feature History for Cisco RF Switch Firmware

Several performance and configuration enhancements have been added to Cisco RF Switch firmware, released in the following most recent versions:

•Version 2.50—SNMPv1 Upconverters and Traps, Default Gateway for Remote TFTP Transfer

•Version 3.30—Improved switchover times, DHCP Server, several new commands or command enhancements for slot configuration and system information

•Version 3.50—Further improved switchover times, optimized ARP cache feature, ARP timeout configuration, and additional show command enhancements for ARP and configuration status

•Version 3.60 includes the following enhancements:

–Changes to the network buffering to allocate a larger pool (number) of buffers, with a new number of 100 buffers total, to help handle an increase in SNMP traffic.

–Reduction of the maximum packet size to 600 bytes. This combination of a larger number of buffers with smaller maximum packet size helps with handling large bursts of inbound packets that were discarded in previous versions of Cisco RF Switch Firmware.

–Resolution of a problem in the SNMP agent to help further with the above items. In prior versions of Cisco RF Switch firmware, the SNMP agent blocked traffic just after packet reception, waiting to allocate a buffer in which to place the output response. If no buffer was available (as would be the case if a large burst of incoming packets occurred), the agent would timeout, and the system would generate a watchdog timeout. Now, the agent uses a private buffer for the output response, and only requests a packet buffer after completing the snmp operation. If no buffer is available, the output response is discarded, and the agent continues processing inbound packets.

–Addition of the noverify option to the copy command, enabling you to override the file type verification, and place a file in either the flash (FL:) or bootflash (BF:) device. Version 3.60 updates the online help to reflect this new option. This new option provides the ability to place a copy of the main application into the bootflash, so that normal system operation is restarted in the case of a system crash, instead of having the "sys>" prompt as in previous versions of Firmware.

–Version 3.60 resolves a previous issue in which concurrent access to the RF switch modules via the command-line interface and SNMP would cause random errors and crashes. The firmware now allows simultaneous usage of telnet, console, and SNMP operation. This issue was observed primarily if the show version and test module commands were used at the same time that SNMP status polling operations were occurring. This previous issue also affected a number of additional commands.

Refer to the Cisco RF Switch Firmware Command Reference Guide on Cisco.com for complete feature descriptions and command histories for the Firmware Versions listed above.

Additional Cisco Broadband Cable High Availability Features

Cisco High Availability (HA) for Broadband Cable products includes these and additional features:

•N+1 HCCP Redundancy

•DOCSIS Stateful Switchover (DSSO)

•Gigabit Ethernet

•PacketCable Support

•Route Processor Redundancy Plus (RPR+)

These and additional HA features are described further in the Cisco White Paper, Cisco Cable IP Solutions for High-Availability Networks, available on Cisco.com.

Finding Support Information for Platforms and Cisco IOS Software Images

Use Cisco Feature Navigator to find information about platform support and Cisco IOS software image support. Access Cisco Feature Navigator at http://www.cisco.com/go/fn. You must have an account on Cisco.com. If you do not have an account or have forgotten your username or password, click Cancel at the login dialog box and follow the instructions that appear.

Contents

This chapter provides the following procedures and commands to configure, test and debug the N+1 Redundancy scheme on your Cisco universal broadband router CMTS:

•Prerequisites

•Restrictions and Limitations

•Information About N+1 Redundancy and the Cisco Universal Broadband CMTS

•Manual RF Switch Configuration Tasks for N+1 Redundancy

•Global N+1 Line Card Redundancy

•How to Configure N+1 Redundancy on the Cisco CMTS

•Switchover Testing Tasks for N+1 Redundancy

•Configuration Examples for Cisco N+1 Redundancy

•Additional References

Prerequisites

To use N+1 HCCP Redundancy, ensure the following conditions are met:

•To implement N+1 Redundancy, you must use an image from a supported Cisco IOS software release. Refer to the release notes for your platform on Cisco.com to verify the availability of the N+1 Redundancy feature.

•Your downstream plant must meet Data-over-Cable Service Interface Specifications (DOCSIS) 1.0 or DOCSIS 1.1 requirements.

•Customer cable modems must meet requirements for your network and server offerings. All third-party cable modems must be DOCSIS 1.0- or DOCSIS 1.1-compliant and configured for two-way data communication.

Restrictions and Limitations

The following sections describe restrictions and guidelines for configuring N+1 line card redundancy.

Note It is important to be aware that in Cisco IOS software releases prior to Cisco IOS Release 12.3(13a)BC, line card redundancy is configured at the interface configuration level using hccp commands. Beginning in Cisco IOS Release 12.3(13a)BC and later, enhancements to the N+1 line card redundancy configuration include a newer command-line interface (CLI) at the global configuration level, that replaces the legacy hccp interface command configuration. The newer feature is referred to as Global N+1 Line Card Redundancy, or Rapid HCCP Configuration. As you consider the restrictions and configuration information in this chapter, keep the distinction between the legacy HCCP configuration and the global configuration in mind.

General N+1 Redundancy Restrictions and Limitations

These restrictions apply to N+1 Redundancy on the Cisco uBR10012 and Cisco uBR7246VXR routers in Cisco IOS Release 12.3(9a)BC and earlier Cisco IOS releases.

•When using the show hccp channel switch Cisco IOS command, the system communicates with each module in the RF Switch that comprises the bitmap. This requires a much longer period for timeout— contrasted with the lesser timeout required for the system to verify connectivity. Use the show hccp g m channel command to view each individual member of an HCCP group.

Cable upstream configuration commands are described in the Cisco Broadband Cable Command Reference Guide on Cisco.com:

http://www.cisco.com/en/US/products/hw/cable/ps2217/products_command_reference_book09186a0080108e88.html

•HCCP interface configuration can be removed from either Working or Protect Interfaces. However, the following HCCP restrictions apply to HCCP N+1 Redundancy on either the Cisco uBR10012 or Cisco uBR7246VXR router:

–Before removing HCCP configurations from an active Working interface, either shut down the Protect or lockout switchover functions using the hccp group lock member-id command in global configuration mode. Otherwise the Protect interface may declare the Working interface to have failed and may attempt to switch over.

–Do not remove HCCP configurations from an active Protect interface. The active HCCP group member should be restored to its corresponding Working interface (revertback) before removing HCCP configuration from the Protect interface.

Note This restriction does not apply when removing HCCP configuration from a Protect interface while it is in standby mode and N+1 Redundancy is in normal Working mode.

For information about modifying HCCP configuration, refer to the section titled "Maintaining Online Cable Modem Service When Removing HCCP Configuration from Working HCCP Interfaces" section.

•Downstream (DS) modulation, interleave depth and DOCSIS Annex mode must be the same for all members in the same HCCP group. For configuration information, refer to the "Preconfiguring HCCP Protect Interfaces for N+1 Redundancy" section.

•When using external, non-SNMP upconverters, DS frequencies must be set to be the same across all cable interface line cards that are protected by the same Protect line card.

N+1 Redundancy Restrictions and Requirements for the Cisco uBR7246VXR Router

Note As of Cisco IOS Release 12.3(21)BC, N+1 redundancy is no longer supported on the Cisco uBR7246VXR router.

•Cisco IOS Release 12.3(17a)BC support 4+1 redundancy on the Cisco uBR7246VXR router with the uBR-MC28C, uBR-MC16S and uBR-MC16C line cards only.

•Global N+1 redundancy configuration is not supported on the Cisco uBR7246VXR router.

•Cisco Systems recommends that the lowest slot interface be the master when configuring cable interface bundling on the Cisco uBR7246VXR router.

•Cisco uBR7246VXR CMTS interfaces that are bundled in IP switch over together.

N+1 Redundancy Restrictions and Requirements for the Cisco uBR10012 Router

Restrictions for Cisco IOS Release 12.2(15)BC2a

If you use DOCSIS 1.1 provisioned cable modems in your network and you are considering deploying Cisco IOS Release 12.2(15)BC2a, Cisco Systems recommends that you disable HCCP N+1 Redundancy until further notification, or that you reduce instances of manual switchover from HCCP Working to Protect via the command line interface (CLI).

Cable interface line cards in HCCP Working or Protect status may reload or experience intermittent failure during HCCP N+1 switchover in Cisco IOS Release 12.2(15)BC2a:

•Cable interface line cards that are in HCCP Working status may reload during N+1 switchover from HCCP Working to Protect status.

•You may experience HCCP memory overrun when cable interface line cards in HCCP Working status switch over to HCCP Protect status.

General Requirements for the Cisco uBR10012 Router with All Cable Interface Line Cards

•A TCC+ card must be installed in your Cisco uBR10012 router in order to employ the Cisco RF Switch in your cable headend system. For more detailed information on the TCC+ card, refer to the Cisco uBR10012 Universal Broadband Router TCC+ Card document available on Cisco.com:

http://www.cisco.com/univercd/cc/td/doc/product/cable/ubr10k/ubr10012/frus/tcc5094.htm

•Use the IP address from the local loopback interface as the Working interface IP address when configuring Hot-Standby Connection-to-Connection Protocol (HCCP) on the Cisco uBR10012 router. Cisco strongly recommends that you create a loopback interface on the Cisco uBR10012 router, and then assign the loopback interface's IP address to the HCCP protect configuration.

•Using slot 5/1 as the Protect interface is easiest for physical wiring to the Cisco RF Switch when used with the Cisco uBR10012 router.

•Cisco IOS downgrade can be performed while retaining N+1 functionality, as supported by earlier Cisco IOS releases. However, when downgrading your Cisco IOS software from release 12.2(15)BC2a to an earlier release, N+1 Redundancy requires that you preconfigure the Protect interface(s) with the cable upstream connector command. Without this HCCP preconfiguration, the upstream channel does not come up again after a switchover.

Note Be careful if you plan to downgrade from Cisco IOS Release 12.3(13a)BC, when the Global N+1 Line Card Redundancy feature was introduced. The global N+1 configuration is not supported in earlier Cisco IOS software releases.

•The HCCP Switchover Enhancements feature in Cisco IOS Release 12.3(21)BC has the following restrictions:

–The feature is supported on the Cisco uBR10012 router with the Cisco Performance Routing Engine 2 (PRE2) only.

–The feature is supported by the following line cards on the Cisco uBR10012 router: Cisco UBR10-MC5X20S, Cisco UBR10-MC5X20U, and Cisco UBR10-MC5X20H

–The line card switchover performance improvements are valid for networks scaling to less than 5000 cable modems per line card, and less than 1000 voice calls per line card.

–The working and protect line cards must have the same channel width.

–Upconverter failure detection is not included as part of the line card switchover performance improvements.

–Virtual interface bundling is required. If you are upgrading from an earlier Cisco IOS software release and virtual bundling is not configured upon startup, the Cisco IOS software will automatically generate a virtual bundling configuration. Therefore, beginning in Cisco IOS Release 12.3(21)BC, Layer 3 information cannot be configured directly at the cable interface. The maximum number of virtual bundle interfaces supported is 40, and bundle numbers can be between 1-255. For more information about configuring virtual interface bundling, see the "Cable Interface Bundling and Virtual Interface Bundling for the Cisco CMTS"chapter.

–Tracking of HCCP interfaces is removed. The hccp track command is obsolete.

–In prior releases, a switchover could be triggered due to a keepalive failure regardless of how many cable modems were online for an upstream. This resulted in false switchovers. In Cisco IOS Release 12.3(21)BC, keepalive failure detection is now enabled only for upstreams that have 15 or greater modems online. However, a switchover due to keepalive failure will trigger only if there is not any traffic on all of the upstreams associated with a cable interface that is enabled for keepalive.

For example, on a cable line card interface enabled for keepalive (this is the default) you have the following US status: US0 (200 CMs online), US1 (10 CMs online), US2 (16 CMs online), US3 (shutdown). US0 and US2 are enabled for keepalive detection because they each have more than 15 modems online.

If US0 has a keepalive failure due to a cable cut, but US2 is still passing traffic, then no keepalive switchover is triggered on that domain or interface. The calculation looks at all relevant US ports in a MAC domain and if those relevant ports have no traffic, then keepalive detection will begin. In this example, only two ports were relevant and both of those ports did not lose traffic, so keepalive still did not activate the failover.

If US0 had a cable cut while US2 also had no traffic, then a keepalive switchover would be triggered.

Restrictions with the Cisco UBR10-MC 5X20U or -S BPE

•MAC domains and corresponding DS interface pairs switch over together. Each ASIC processor on the Cisco UBR10-MC 5X20U or -S BPE supports two MAC domains. MAC domains that share a common ASIC processor (JIB) must be configured so that they share the same state, Active or Standby. As a result, each interface in the pair switches over with the other.

Downstream MAC domain pairings would be downstream (DS) ports 0 and 1, ports 2 and 3, and a solitary port 4, which has its own JIB. For example, these interface pairings share the same JIB and switch over together as follows:

–Cable interface 5/0/0 and 5/0/1

–Cable interface 5/0/2 and 5/0/3

–Cable interface 5/0/4 is on the third ASIC processor, which is not shared with another interface.

Note If HCCP is not configured on an interface that shares a MAC processor with another configured interface, it does not switch over and could cause issues. The same holds true if an ASIC companion is "locked out" during a failover.

Disabling HCCP Revertive on Protect Cable Interfaces

The cable interface line cards pair up interfaces that share the same JIB (ASIC processor) as explained in the restriction immediately above.

As a result, when HCCP keepalive is enabled on paired DS channels, both DS channels in the pair switch over together if either DS channel has a keepalive failure. For example, if HCCP is configured on DS channels 0 and 1, and DS channel 0 has a keepalive failure, then DS channel 1 also fails because it shares the same JIB with DS channel 0.

When HCCP revertive is enabled on both downstream channels in the pair, the interface that experiences the keepalive failure does not revert back automatically to active state. This is desirable behavior because it prevents revertback to active state prematurely—before the cause of an external failure is confirmed and remedied.

Note The default HCCP revertive time for HCCP interfaces is 30 minutes.

However, the JIB companion interface may act upon the default revertive time of 30 minutes. The companion interface attempts to revert back to active state after 30 minutes (when HCCP revertive is enabled). This creates conflict with the failed companion interface on the same JIB.

Note Therefore, Cisco Systems recommends that you disable automatic HCCP revertive functions on both Protect downstream channels of a JIB that use keepalive or tracking. If you have keepalive and tracking enabled, or you are using the UBR10-MC 5X20U or -S in N+1 configuration, disable the revertive function on both Protect interfaces.

To disable the HCCP revertive function on Protect interfaces, use the no hccp group revertive command in cable interface configuration mode. Disable revertive on each HCCP Protect interface:

no hccp group revertive

Syntax Description

group

The group number for the specified interface. Valid values are any number from 1 to 255, inclusive.

For additional information about configuring or removing HCCP, refer to the "How to Configure N+1 Redundancy on the Cisco CMTS" section, and to the hccp revertive command in the Cisco Broadband Cable Command Reference Guide on Cisco.com:

http://www.cisco.com/univercd/cc/td/doc/product/cable/bbccmref/index.htm

Information About N+1 Redundancy and the Cisco Universal Broadband CMTS

This section describes the following concepts that relate to N+1 Redundancy:

•The Components and Terminology of N+1 Redundancy

•IF Muting on the Cisco CMTS for non-SNMP-capable Upconverters

•DSX Messages and Synchronized PHS Information

•High Availability Support for Encrypted IP Multicast

The Components and Terminology of N+1 Redundancy

N+1 Redundancy is made possible with the addition of the Cisco RF Switch to your cable headend network. The N+1 Redundancy protection scheme you select for your system depends on your CMTS platform and upon the number of cable interface line cards or Broadband Processing Engines (BPEs) that you have installed in the Cisco router chassis.

N+1 Redundancy is available for these Cisco Cable Modem Termination System (CMTS) platforms:

Table 1 Cisco CMTS Platforms Supporting N+1 Redundancy

CMTS Platform/N+1

Line Cards or BPEs

Supported Upconverters

Cisco RF Switch

Cisco uBR10012

•UBR10-LCP2-MC16C

•UBR10-LCP2-MC16C=

•UBR10-LCP2-MC16E

•UBR10-LCP2-MC16E=

•UBR10-LCP2-MC16S

•UBR10-LCP2-MC16S=

•UBR10-LCP2-MC28C

•UBR10-LCP2-MC28C

•UBR10-MC5X20U, -S, or -H

•SNMP with RF Muting

•Non-SNMP¹ with IF Muting

Cisco 3x10 RF Switch (one or multiple)

Cisco uBR7246VXR

•UBR-MC28C

•UBR-MC16S

•UBR-MC16C

•SNMP with RF Muting

•Non-SNMP1 with IF Muting

Cisco 3x10 RF Switch (two)

¹Non-SNMP upconverters are supported beginning with Cisco IOS Release 12.2(15)BC2a.

N+1 Redundancy refers to Working cable interface line cards (N) being protected by one additional line card (+1). The two types of Cisco N+1 configuration are as follows:

•8+1 (7+1)—Refers to an eight-card redundancy scheme in which seven Working cable interface line cards are protected by one additional Protect line card. This is the default N+1 configuration for the Cisco uBR10012 router. This redundancy scheme is also referred to as 7+1 redundancy, which is the more physically accurate term.

•4+1—Refers to a four-card redundancy scheme in which four Working cable interface line cards are protected by one additional Protect line card.

Upconverters may reside between the Cisco RF Switch and the downstream (DS) interface on the Cisco CMTS. Cisco IOS supports both SNMP and non-SNMP-capable upconverters.

N+1 Redundancy on the Cisco uBR10012 Universal Broadband Router

The eight-card 7+1 Redundancy scheme for the Cisco uBR10012 router supports redundancy for the cable interface line cards installed in a fully populated Cisco uBR10012 chassis. Other redundancy schemes are designed to support partial cable interface line card populations in a Cisco uBR10012 chassis.

A single Cisco uBR10012 CMTS can support up to eight Cisco cable interface line cards, each featuring one to five downstream and six to 20 upstream cable interfaces for a total of up to 40 downstream and 160 upstream interfaces in the chassis.

A single Cisco RF Switch can then be connected to this Cisco uBR10012 CMTS, allowing you to deploy an N+1 Redundancy scheme where one protecting cable interface line card supports from one to seven Working cable interface line cards in the same chassis.

The Cisco uBR10012 router supports N+1 Redundancy on the following Cisco uBR10012 cable interface line cards (broadband processing engines—BPEs):

Cable Interface Line Card

N+1 Redundancy Introduced

Cisco UBR10-MC5X20H

Cisco IOS Release 12.3(17a)BC2

Cisco UBR10-MC 5X20U or -S

Cisco IOS Release 12.2(15)BC2a

Cisco UBR10-MC 5X20U or -S

Cisco IOS Release 12.2(15)BC1

Cisco uBR10-LCP2-MC16C,
Cisco uBR10-LCP2-MC16E,
Cisco uBR10-LCP2-MC16S

Cisco IOS Release 12.2(8)BC2

Note Beginning in Cisco IOS Release 12.2(15)BC2a, these cable line card interfaces are end-of life (EOL).

UBR10-LCP2-MC28C

Cisco IOS Release12.2(4)XF1, 12.2(4)BC1

The Cisco uBR10012 router contains eight slots, numerated as shown in Figure 1, using the slot/port CLI convention (for example, slot 8/0).

A Cisco uBR10012 router identifies a subinterface addresses by slot number, subslot number, and downstream (DS) port number, in the format slot/subslot/DS port. For example, the address of a subinterface could be 5/1/0 (slot 5, subslot 1 and DS port 0).

Cisco IOS command line syntax is unique when selecting or defining slots, subslots and ports for the Cisco uBR10012 router. For example, the syntax of the Cisco IOS command
interface cable slot/subslot/port identifies a cable interface on the Cisco uBR10012 router. The following are the valid values for this and similar such commands:

•slot = 5 to 8

•subslot = 0 or 1

•port = 0 to 4 (depending on the cable interface)

Figure 1 illustrates the numeration of these cable interfaces on the Cisco uBR10012 router chassis.

Chassis Slot Numeration and Selection on the Cisco uBR10012 Router

Figure 1 Cisco uBR10012 Chassis Slot Numbering —Rear View

N+1 Redundancy on the Cisco uBR7246VXR Universal Broadband Router

The 4+1 redundancy scheme for the Cisco uBR7246VXR router supports redundancy for the cable interface line cards installed in four fully populated router chassis.

Note Cisco Systems recommends using the chassis with the most memory, network processing engine (NPE) power and additional resources as the Protect chassis.

Each Cisco uBR7246VXR can support up to four Cisco cable interface line cards, each featuring one or two downstream and six or eight upstream cable interfaces, for a total of up to eight downstream and 32 upstream interfaces in the chassis.

Two Cisco RF Switches can be connected to four Working and one Protect Cisco uBR7246VXR routers, allowing you to deploy an N+1 Redundancy scheme in which one protecting cable interface line card in the Working uBR7246VXR supports one Working cable interface line card in each of the four Working chassis.

The Cisco uBR7246VXR router supports N+1 Redundancy on the following cable interface line cards:

Cable Interface Line Card

N+1 Redundancy Introduced

Cisco uBR-MC16S/C

Cisco IOS Release 12.2(15)BC2a

Cisco uBR-MC28C

Cisco IOS Release 12.2(15)BC2a

Chassis Slot Numeration on the Cisco uBR7246VXR Router

For Cisco uBR7200 series components, the slot number is the chassis slot in which a port adapter or a cable interface card is installed. The logical interface number is the physical location of the interface port on a port adapter. Numbers on a Cisco uBR7200 series router begin with 0.

Using a Cisco uBR7246VXR router chassis to illustrate, slot/port positioning is as follows:

•Slot 0—I/O controller

•Slot 1-2—Cisco port adapters

•Slot 3-6—Cisco cable interface line cards; the upstream ports on the card start with port 0.

For the Cisco uBR7246VXR reference design discussed in this guide, line card (LC) 1 in Cisco uBR7246VXR 5 protects the Working LC 1 in router chassis 1, 2, 3, and 4. LC 2 in chassis 5 protects the Working line card 2 in chassis 1, 2, 3, and 4, and so forth.

Figure 2 Cisco uBR7246VXR Router Chassis Slot Numbering—Rear View

N+1 Redundancy and the Cisco RF Switches

The Cisco RF Switch can be operated in two separate modes, either in 8+1 configuration, or in 4+1 configuration as two RF Switches.

Note The default N+1 Redundancy mode for the Cisco RF Switch is 8+1. This does not require change when configuring N+1 Redundancy on the Cisco uBR10012 router with the Cisco UBR10-MC 5X20U or -S BPE.

Note The show configuration command and other Cisco RF Switch commands contain the Card Protect Mode field. When this field displays 8+1, this indicates that the Cisco RF Switch in configured for N+1 Redundancy, where eight or less Working line cards are possible.

Cisco 3x10 RF Switch Chassis Overview

Figure 3

Cisco RF Switch Chassis—Front View

In both of the Cisco RF Switches, the slot number is the chassis slot in which an Ethernet controller or an upstream or downstream card is installed, and the logical interface number is the physical location of the interface port on an Ethernet controller.

The MAC-layer or hardware address is a standardized data link layer address that is required for certain network interface types. The Cisco RF Switch uses a specific method to assign and control the MAC-layer addresses of its Ethernet controller.

The Ethernet controller and upstream and downstream assembly slots maintain the same slot number regardless of whether other Ethernet controllers or upstream or downstream cards have been installed or removed. However, when you move an upstream or downstream card to a different slot, the logical interface number changes to reflect the new slot number. The Ethernet card is always installed in the same slot.

All LAN interfaces (ports) require unique MAC-layer addresses, also known as hardware addresses. Typically, the MAC address of an interface is stored on a memory component that resides directly on the interface circuitry; however, the OIR feature requires a different method.

The OIR feature allows you to remove an Ethernet controller or an upstream or downstream assembly and replace it with another identically configured one. If the new controller or assembly matches the controller or assembly you removed, the system immediately brings it online. In order to allow OIR, an address allocator with a unique MAC address is stored in an EEPROM on the Cisco RF Switch midplane. Each address is reserved for a specific port and slot in the switch, regardless of whether an Ethernet controller or an upstream or downstream assembly resides in that slot.

The MAC addresses are assigned to the slots in sequence. The first address is assigned to Ethernet controller slot 0, and the next addresses are assigned to upstream and downstream assembly slots 1 through 14. This address scheme allows you to remove the Ethernet controllers or assemblies and insert them into other switches without causing the MAC addresses to move around the network or be assigned to multiple devices.

Cisco RF Switch Modules

Figure 4 Cisco RF Switch Modules, Rear View

The Cisco RF switch module is a switching matrix that allows flexibility in the routing of RF signals between "N" Working RF cable interface line cards and one Protect RF cable interface line card.

The RF Switch header block has 14 ports labeled with letters. Each header screws into a slot in the Cisco RF Switch. A Cisco RF Switch module contains all the active relays for a particular port for all slots.

Cisco uBR 3x10 RF Switch Slot Information

Table 2 lists the RF modules and the ports assigned to each module, as illustrated in Figure 4.

Tip The modules are listed as seen from the front of the RF switch.

Table 2 Switching Matrix for the Cisco uBR 3x10 RF Switch (Upstream and Downstream Modules)

RFS Module

Working Ports

PROTECT Ports

Type

RFS Module

Working Ports

PROTECT Ports

Type

2

1H—8H

P1H, P2H¹

upstream

1

1A—8A

P1A, P2A

upstream

4

1I—8I

P1I, P2I

upstream

3

1B—8B

P1B, P2B

upstream

6

1J—8J

P1J, P2J

upstream

5

1C—8C

P1C, P2C

upstream

8

1K—8K

P1K, P2K

upstream

7

1D—8D

P1D, P2D

upstream

10

1L—8L

P1L, P2L

upstream

9

1E—8E

P1E, P2E

upstream

12

1M—8M

P1M, P2M

downstream

11

1F—8F

P1F, P2F

downstream

14

not used

—

—

13

1G—8G

P1G, P2G

downstream

¹P2 is used only when the switch is in 4 + 1 mode.

8

Example:

Modules 1-10 below are upstream (US) modules in the Cisco uBR 3x10 RF Switch.

The remainder of the modules are either assigned to downstream functions or are not used.

•Module 1 uses Port a for slots 1-8 on the Working, and it uses Port a of Protect slot 1 and/or Protect slot 2.

•Module 2uses CMTS Ports 1h through 8h, and Protect Port 1h and Protect Port 2h.

•Module 3 uses port b.

•Module 4 uses port i.

•Module 5 uses port c.

•Module 6 uses port j.

•Module 7 uses port d.

•Module 8 uses port k.

•Module 9 uses port e.

•Module 10 uses port l.

•Module 11 uses port f.

•Module 12 uses port m.

•Module 13 uses port g.

•Module 14 uses port n, which is not used on the Cisco uBR 3x10 RF Switch.

The Cisco uBR 3x10 RF Switch works with the Cisco uBR10012 router and supports three downstream modules and 10 upstream modules. Each RF switch module supports the full frequency range specified by DOCSIS and EuroDOCSIS standards.

IF Muting on the Cisco CMTS for non-SNMP-capable Upconverters

Beginning with Cisco IOS Release 12.2(15)BC2a, Cisco supports IF Muting with both SNMP and non-SNMP-capable upconverters in N+1 Redundancy. IF Muting offers the following benefits:

•IF Muting for either type of upconverter significantly increases the N+1 protection schemes that are available for Cisco CMTS headends.

•IF Muting offers the additional benefit of being faster than RF Muting.

•IF Muting is enabled by default. The Cisco CMTS automatically enjoys the benefits and availability of IF Muting.

IF Muting functions in the following manner:

•IF output from the Working cable interface line card is enabled.

•IF output from the Protect cable interface line card is disabled.

•When a switchover occurs from Working to Protect, the IF output of the Working card is disabled and that of the Protect is enabled. If an interface is in Active mode, RF output is enabled.

•When the cable interface line card first comes up after a system failure, IF output is muted until the Cisco CMTS determines if each interface is in active or standby mode (in either Working or Protect state). When an interface is active (Working or Protect), IF output is enabled. When an interface is in standby mode, IF output is muted.

The relevance and support for IF Muting is dependent on the type of Cisco CMTS being used. This is a summary of IF Muting in relation to three sample scenarios:

•Case1—External upconverters are not controlled nor controllable. In this type of scenario, the external upconverter either cannot be controlled remotely or the Cisco CMTS is not configured to control the external upconverter.

•This type of Cisco CMTS is newly supported with Cisco IOS Release 12.2(15)BC2a. Previously, such customers could not enable N+1 Redundancy in the Cisco CMTS headend because they use upconverters that previously could not be controlled from the Cisco CMTS.

•Case 2—The Cisco CMTS is configured to control an external upconverter. Cisco continues to support N+1 Redundancy in this scenario (in which IF Muting is not required). The Cisco CMTS uses RF Muting of the upconverter in this scenario—automatically enabled when an HCCP upconverter statement is configured.

•Case 3—The Cisco CMTS uses internal upconverter(s), as with the Cisco UBR10-MC 5X20U or -S BPE. Cisco continues to support N+1 Redundancy in this scenario (in which IF muting is not required). The Cisco CMTS uses RF muting in this scenario (automatically enabled) because the upconverter is configured by the CMTS to do RF Muting.

IF Muting and HCCP Configuration

HCCP interface configuration typically entails three tasks:

•Working or Protect mode

•Upconverter statement

•RF switch statement

When you configure HCCP on an interface, but you do not specify an upconverter statement, this dictates whether IF Muting is active. With no upconverter statement in the interface configuration, IF Muting becomes active by default.

For additional details, refer to the procedures in these sections:

•Manual RF Switch Configuration Tasks for N+1 Redundancy

•How to Configure N+1 Redundancy on the Cisco CMTS

Restrictions for IF Muting

Shared Downstream Frequency

All the interfaces in the same HCCP group must use the same downstream frequency. To define the downstream center frequency for the cable interface line card, use the cable downstream frequency command in cable interface configuration mode. On cable interfaces with an integrated upconverter, use the no form of this command to remove the downstream frequency and to disable the RF output.

cable downstream frequency down-freq-hz

no cable downstream frequency

The no form of this command is supported only on the Cisco uBR-MC28U/X cable interface line card and the UBR10-MC 5X20U or -S.

•down-freq-hz—The known center frequency of the downstream carrier in Hz (the valid range is 55 MHz to 858 MHz). The usable center frequency range depends on whether the downstream is configured for DOCSIS or EuroDOCSIS operations:

–DOCSIS — 91 to 857 MHz

–EuroDOCSIS — 112 to 858 MHz

The Cisco IOS supports a superset of these standards, and setting a center frequency to a value outside these limits violates the DOCSIS or EuroDOCSIS standards. Cisco does not guarantee the conformance of the downstream and upconverter outputs when using frequencies outside the DOCSIS or EuroDOCSIS standards.

For additional information about this command, refer to the Cisco Broadband Cable Command Reference Guide on Cisco.com.

Requirements for IF Muting

For non-SNMP-capable upconverters to be used with IF Muting, RF output must be less than -3 dBmV when:

•IF input is absent.

•The switchover time from Working to Protect is less than one second. That is, when IF is applied to the upconverter, the RF output must be present within one second.

If either of these requirements is not met, the integrity of the N+1 switchover operations could be compromised.

DSX Messages and Synchronized PHS Information

Cisco IOS Release 12.3(17a)BC introduces support for PHS rules in a High Availability environment. In this release, and later releases, PHS rules synchronize and are supported during a switchover event of these types:

•Route Processor Redundancy Plus (RPR+) for the Cisco uBR10012 router, with Active and Standby Performance Routing Engines (PREs)

•HCCP N+1 Redundancy, with Working and Protect cable interface line cards

For further information about DSX messages and Payload Header Suppression (PHS) information on the Cisco CMTS, refer to these documents, and additional DOCSIS PHS information:

•Cable DOCSIS 1.1 FAQs, Cisco TAC Document 12182

http://www.cisco.com/en/US/tech/tk86/tk168/technologies_q_and_a_item09186a0080174789.shtml

•DOCSIS 1.1 for the Cisco CMTS

http://www.cisco.com/en/US/products/hw/cable/ps2217/products_feature_guide_chapter09186a008019b57f.html

High Availability Support for Encrypted IP Multicast

Cisco IOS Release 12.3(17a)BC introduces support for IP Multicast streams during switchover events in a High Availability environment. This feature is supported for Route Processor Redundancy Plus (RPR+), N+1 Redundancy, and encrypted BPI+ streams.

For additional information about IP Multicast and High Availability, refer to these documents on Cisco.com:

•Cisco CMTS Universal Broadband Router MIB Specifications Guide

http://www.cisco.com/en/US/products/hw/cable/ps2217/products_technical_reference_chapter09186a00805fd8fb.html

•Dynamic Shared Secret for the Cisco CMTS

http://www.cisco.com/en/US/products/hw/cable/ps2217/products_feature_guide09186a00801b17cc.html

•IP Multicast in Cable Networks, White Paper

http://www.cisco.com/en/US/tech/tk828/technologies_case_study0900aecd802e2ce2.shtml

•Route Processor Redundancy Plus for the Cisco uBR10012 Router

http://www.cisco.com/en/US/products/hw/cable/ps2209/products_feature_guide09186a00801a24e0.html

Manual RF Switch Configuration Tasks for N+1 Redundancy

You must configure and activate both the Cisco RF Switch and the Cisco CMTS to ensure that
N+1 Redundancy operates correctly. You must also configure HCCP Working interfaces and groups.

Perform these procedures in sequence when configuring N+1 Redundancy on the Cisco RF Switch.

Procedure

Purpose

Step 1

"Configuring the Cisco RF Switch for N+1 Redundancy" procedure

(Required) Provides required and optional configurations on the Cisco RF Switch, including MAC and IP addressing, SNMP configurations, and switchover interface groups.

Step 2

"Creating Cisco RF Switch Module Bitmaps" procedure

(Required) Provides required configuration of hexadecimal-format module bitmaps that indicate which upstream (US) and downstream (DS) modules belong to a switchover group.

Configuring the Cisco RF Switch for N+1 Redundancy

SUMMARY STEPS

1. set mac address mac-address (optional)

2. set ip address ip-address netmask (optional)

3. set slot config {upstreamslots | downstreamslots } (optional)

4. set snmp community read-write private (optional)

5. set snmp host ip-address (optional)

6. set snmp traps (optional)

7. set protection {4|8} (required)

8. set password text (optional)

9. set tftp-host ip-address (optional)

10. set switchover-group group-name module-bitmap | all (required)

DETAILED STEPS
Command

Purpose
Step 1
set mac address mac-address
Example:

rfswitch> set mac address 0000.8c01.1111
(Optional) To specify the MAC address of the Ethernet port on the Cisco RF Switch (used to connect to the LAN), use the set mac address command at the Cisco RF Switch command line interface.

The MAC address must be specified using a trio of hexadecimal values. For example, set mac address hex.hex.hex. To negate the existing MAC address assignment and specify a new one, use the no form of this command. If no MAC address is specified, the Cisco RF Switch assumes the default OUI MAC address value.
Step 2
set ip address ip-address 
netmask [ dhcp ]
Example:

rfswitch> set ip address 172.16.10.3 255.255.255.0
(Optional) To specify a static IP address and relative netmask of the Ethernet interface on the Cisco RF Switch, use the set ip address command in User mode. To restore the default setting, user the no form of this command.

Default setting differs according to your Firmware Version:

•The default IP configuration for Version 3.30 and 3.50 is DHCP enabled.

•The dhcp keyword enables the specified IP address as the address for DHCP services on the network. This keyword also produces the same result as the no form of this command for Version 3.30 and 3.50—it enables DHCP.

•The default IP configuration for Version 2.50 is the static IP address of 10.0.0.1 255.255.255.0.
Step 3
set slot config {upstreamslots 
| downstreamslots }
Example:

Cisco 3x10 RF Switch (default)
rfswitch> set slot config 
0x03ff 0x1c00
(Optional) Sets the chassis slot-to-line card configuration. The command no set slot config restores the default, which is a 3x10 configuration.

Setting a bit position tells the Cisco RF Switch to expect that type of card installed in the slot. A zero in both parameters indicates that the slot should be empty. Both upstreamslots and dnstreamslots are 16-bit hex integer bit-masks that represent whether the slot is enabled/configured for that type of card. The right-most bit represents slot 1.

For additional bitmap conversion information, refer to the Bitmap Calculator for N+1 Configuration with the Cisco RF Switch (Microsoft Excel format)

http://www.cisco.com/warp/public/109/BitMap.xls

As there are only 14 slots in the Cisco RF Switch ch