Table Of Contents
Performance Route Processor (PRP) Software Configuration in Cisco 12000 Series Routers
This feature was introduced on Cisco 12000 Series Routers.
Disabling of Layer 3 cache added.
This feature module describes the software configuration of the Performance Route Processor (PRP) in Cisco 12000 Series Routers.
This document includes the following sections:
The Performance Route Processor (PRP) in the Cisco 12000 Series Router is designed to meet the growing demands in the high-end IP routing market due to increased Internet traffic, larger routing tables, acceptance of IP as the common networking layer protocol, and an increasing shift in focus by service providers to IP infrastructures.
The PRP satisfies these demands by providing an evolution from the current Gigabit Route Processor (GRP) card with a faster processor, multilayer cache, improved fabric interface for faster communication between the PRP and line cards, and larger memory capacity.
The PRP is designed to support all IOS software features that are supported on the GRP, such as High Availability and standard routing protocols.
Note The Stateful Switchover with Nonstop Forwarding (SSO/NSF) feature will be supported in IOS Release 12.0(23)S and later.
Route Processor Redundancy Plus (RPR+) Support
The PRP supports the High Availability feature, Route Processor Redundancy Plus (RPR+). With RPR+, the standby RP in a Cisco 12000 Series Router is fully initialized and configured.
RPR+ dramatically shortens the switchover time if the active RP fails or if a manual switchover is performed. Because both the startup configuration and running configuration are continually synchronized from the active to the standby RP, line cards are not reset during a switchover. The interfaces remain up during this transfer, so neighboring routers do not detect a link flap (the link does not go down and back up). For more information, refer to:
The exception warmstart feature in PRP software minimizes downtime following a system crash. It allows for a faster recovery by avoiding the need to reload the software image from an external device.
PRP and GRP Coexistence
You cannot use the PRP and the GRP at the same time during normal router operation. The only time when a GRP and PRP can be used in the same router is during migration from the GRP to the PRP. For more information about the migration procedure, refer to the Performance Route Processor Installation and Configuration Note.
Supported Software Images
You cannot load a GRP software image (with the naming convention gsr-p-mz) on the PRP. GRP and PRP software images are not interchangeable due to the different types of CPU used in each route processor. The PRP uses a software image with the naming convention c12kprp-p-mz.
Redundant PRP Installations
The Warmstart feature is disabled by default if a second, backup PRP is installed in the router for redundancy.
•Cisco IOS Configuration Fundamentals Configuration Guide, Release 12.0
•Cisco IOS Configuration Fundamentals Command Reference, Release 12.0
•Cisco 12000 Series Internet Router Flash Disk Information (DOC-7814024)
•Cisco 12000 Series Internet Router Memory Replacement Instructions (DOC-784338)
•Performance Route Processor Installation and Configuration Note (DOC-7813302)
•Route Processor Redundancy Plus for the Cisco 12000 Series Internet Router
•Performance Route Processor Installation and Configuration (DOC-7813302)
•Cisco 12000 Series Routers
Determining Platform Support Through Cisco Feature Navigator
Cisco IOS software is packaged in feature sets that support specific platforms. To get updated information regarding platform support for this feature, access Cisco Feature Navigator. Cisco Feature Navigator dynamically updates the list of supported platforms as new platform support is added for the feature.
Cisco Feature Navigator is a web-based tool that enables you to determine which Cisco IOS software images support a specific set of features and which features are supported in a specific Cisco IOS image. You can search by feature or release. Under the release section, you can compare releases side by side to display both the features unique to each software release and the features in common.
To access Cisco Feature Navigator, you must have an account on Cisco.com. If you have forgotten or lost your account information, send a blank e-mail to firstname.lastname@example.org. An automatic check will verify that your e-mail address is registered with Cisco.com. If the check is successful, account details with a new random password will be e-mailed to you. Qualified users can establish an account on Cisco.com by following the directions at http://www.cisco.com/register.
Cisco Feature Navigator is updated regularly when major Cisco IOS software releases and technology releases occur. For the most current information, go to the Cisco Feature Navigator home page at the following URL:
Availability of Cisco IOS Software Images
Platform support for particular Cisco IOS software releases is dependent on the availability of the software images for those platforms. Software images for some platforms may be deferred, delayed, or changed without prior notice. For updated information about platform support and availability of software images for each Cisco IOS software release, refer to the online release notes or, if supported, Cisco Feature Navigator.
Supported Standards, MIBs, and RFCs
No new or modified standards are supported by this feature.
No new or modified MIBs are supported by this feature.
To obtain lists of supported MIBs by platform and Cisco IOS release, and to download MIB modules, go to the Cisco MIB website on Cisco.com at the following URL:
No new or modified RFCs are supported by this feature.
See the following sections for configuration tasks for the Warmstart feature. Each task in the list is identified as either required or optional.
•Configuring Warmstart (required)
•Verifying Warmstart (optional)
You can reconfigure the Warmstart feature only in a single PRP configuration. In a dual (redundant) PRP configuration, the Warmstart feature is disabled by default and should not be enabled.
To configure the Warmstart feature, use the following commands starting in global configuration mode:
To verify the configuration settings for the warmstart feature, use the show run command in global configuration mode. In the following example, the warmstart configuration settings are shown in bold.Router# show run...line con 0exec-timeout 0 0line aux 0line vty 0 4login...exception warmstart 60 5end
Monitoring and Maintaining the PRP Software Configuration
To monitor the software configuration of the PRP, use the following show commands in EXEC mode:
This following example shows how to configure the warmstart feature on a Cisco 12000 Series Router so that the IOS software restarts after a crash if the PRP in the router has already been running for at least 60 seconds and if the IOS software has not been restarted by the Warmstart feature more than 10 times:Router# configure terminalRouter# exception warmstart 60 10Router# end
This section documents new and modified commands. All other commands used with this feature are documented in the Cisco IOS Release 12.0 command reference publications.
cache l3 disable
To disable the Layer 3 cache on the route processor (RP), use the cache l3 disable command in global configuration mode. To reenable the Layer 3 cache, use the no form of this command.
cache l3 disable
no cache l3 disable
This command has no arguments or keywords.
Use this command in instances where cache parity errors are of issue. Disabling the Layer 3 cache on the RP will minimize the chance of soft errors in the CPU.
To reset and restart all packet statistics maintained in the PRP segmentation and reassembly (PSAR) drivers, use the clear psar command in global configuration mode.
This command has no arguments or keywords.
No default behavior or values.
Global configuration mode.
The clear psar command replaces the clear csar command used on the Gigabit Route Processor.
The following example shows how to clear all statistics recorded in the PSAR drivers:Router# clear psar
To configure a Cisco 12000 Series Router for a warmstart in case of a system crash, use the exception warmstart command in global configuration mode. To remove the warmstart configuration settings, use the no form of this command.
exception warmstart min-uptime max-restarts [d]
no exception warmstart min-uptime max-restarts
The default for the minimum uptime is 60 seconds.
The default for the maximum number of restarts allowed is 5.
By default, the warmstart feature is:
•Enabled if only one PRP is installed.
•Disabled if a second, standby PRP is installed in the router.
Global configuration mode.
The warmstart feature allows the PRP in a Cisco 12000 Series Router to restart the IOS software configuration after a crash, without having to reload the image from an external device.
Note As with other exception commands, use the exception warmstart command only as instructed and when asked to do so by Cisco technical support personnel.
If the PRP has been running for at least the amount of time specified by min-uptime, and if the system has not been restarted more than the number of times specified by max-restarts, the route processor restarts IOS following a system crash.
In a dual PRP configuration (that is, when a redundant PRP is installed in the router), the warmstart feature is disabled by default. For this reason, you must specify d (for dual) when you enter the exception warmstart command to enable a warmstart.
This feature does not affect the behavior of the reload command. Also, you can still perform a "send break" as usual from the console by pressing Ctrl-Z.
The following example applies to a redundant configuration in which two PRPs are installed. It shows how to configure a warmstart if the IOS software in the PRP has been running for at least 90 seconds, and if the system has not been restarted more than eight times:Router# configure terminalRouter# exception warmstart 90 8 d
show controllers ethernet
To display the register values and status of the GT64260 Ethernet controllers, use the show controllers ethernet command in privileged EXEC mode.
show controllers ethernet [mib | registers | brief]
No default behavior or values.
Use the show controllers ethernet command for debugging purposes to provide troubleshooting information for a Cisco technical support engineer.
The following example shows how to display MIB statistics for the ethernet0 and ethernet1 management interfaces:Router# show controllers ethernet mib---------------------------------------------------Ethernet0 - GT64260 10/100 Mbps Ethernet ControllerMIB Counters:bytes_received = 0bytes_sent = 0frames_received = 0frames_sent = 0total_bytes_received = 0total_frames_received = 0broadcast_frames_received = 0multicast_frames_received = 0crc_error = 0oversize_frames = 0fragments = 0jabber = 0collision = 0late_collision = 0frames_64_bytes = 0frames_65_to_127_bytes = 0frames_128_to_255_bytes = 0frames_256_to_511_bytes = 0frames_512_to_1023_bytes = 0frames_1024_to_max_bytes = 0rx_error = 0dropped_frames = 0out_multicast_frames = 0out_broadcast_frames = 0out_unicast_frames = 0undersize_frames = 0---------------------------------------------------Ethernet1 - GT64260 10/100 Mbps Ethernet ControllerMIB Counters:bytes_received = 0bytes_sent = 0frames_received = 0frames_sent = 0total_bytes_received = 0total_frames_received = 0broadcast_frames_received = 0multicast_frames_received = 0crc_error = 0oversize_frames = 0fragments = 0jabber = 0collision = 0late_collision = 0frames_64_bytes = 0frames_65_to_127_bytes = 0frames_128_to_255_bytes = 0frames_256_to_511_bytes = 0frames_512_to_1023_bytes = 0frames_1024_to_max_bytes = 0rx_error = 0dropped_frames = 0out_multicast_frames = 0out_broadcast_frames = 0out_unicast_frames = 0undersize_frames = 0
show controllers gt64260
To display the register values of the GT64260 Discovery System controller, use the show controllers gt64260 command in privileged EXEC mode.
show controllers gt64260 [CPU | IDMA]
Displays current register values of the Discovery CPU interface.
Displays current register values of the Discovery IDMA interface.
No default behavior or values.
Use the show controllers gt64260 command for debugging purposes to provide troubleshooting information for a Cisco technical support engineer.
The following example shows how to display the register values of the Discovery CPU interface:Router# show controllers gt64260 cpuGT64260 CPU Interface registers:Access Protection Regs -Chan 0 Lo 00000FFF Hi 00000000Chan 1 Lo 00000FFF Hi 00000000Chan 2 Lo 00000FFF Hi 00000000Chan 3 Lo 00000FFF Hi 00000000Chan 4 Lo 00000FFF Hi 00000000Chan 5 Lo 00000FFF Hi 00000000Chan 6 Lo 00000FFF Hi 00000000Chan 7 Lo 00000FFF Hi 00000000
show controllers psar
To display statistics about the packets sent and received in the PRP packet segmentation and reassembly (PSAR) controllers, use the show controllers psar command in privileged EXEC mode.
show controllers psar [drop | fromfab | output | queue | register | tofab]
No default behavior or values.
Use the show controllers psar command for debugging purposes to provide troubleshooting information for a Cisco technical support engineer.
The show controllers psar command replaces the show controllers csar command used on the Gigabit Route Processor.
The following example shows how to display the number of packets sent to the fabric interface. Note that this command only shows statistics to a given slot that have non-zero statistics. In this example, packets are sent only to slots 2, 4 and 5.Router# show controllers psar tofabTo Fabric Stats:---------------Slot Tx pkts Tx TH pkts Tx dropped Tx DMA2 0 2407 0 24074 0 43 0 435 12 2448 0 2460To Fabric Errors:----------------Failed sends because of no header bufs - 0Failed sends because of no payload bufs - 0SRAM parity errors - 0DMA errors - 0
The next example shows how to display the number of packets dropped by the PSAR driver because of lack of buffers, packet headers, no room in the descriptor queues, and so on:Router# show controllers psar dropTo/From Fabric dropped packets:------------------------------Slot Tx pkt Rx pkt0 0 01 0 02 0 03 0 04 0 05 0 06 0 07 0 08 0 09 0 010 0 011 0 012 0 013 0 014 0 015 0 0mcast 0 0
The next example shows how to display statistics for each PSAR queue, including the number of packet headers available in the PSAR driver pool:Router# show controllers psar queueTo Fabric Queue Stats:---------------------Free packet header buffers 9216Slot Cur Entrys Max Entrys Blocked BP count0 0 0 No 01 0 0 No 02 0 4 No 03 0 0 No 04 0 2 No 05 0 4 No 06 0 0 No 07 0 0 No 08 0 0 No 09 0 0 No 010 0 0 No 011 0 0 No 012 0 0 No 013 0 0 No 014 0 0 No 015 0 0 No 0mcast 0 0 No 0
The following example shows how to display information about the PSAR driver register:Router# show controllers psar registerChopper Registers:-----------------FPGA Version 00000095Descriptor fetch enable 00000000Descriptor array size 000003FFInterrupt cause 00000000Interrupt mask FC01FFFFDiscovery low control reg base address F1000840Discovery next desc reg base address F1000830Chopper PCI base address F0000000Descriptor queue empty status 0001FFFFDescriptor queue almost full 00000000Data queue empty status 0001FFFFData queue almost full 00000000SRAM Descriptor threshold 000003F0SRAM Data threshold 000007F0DMA Control 000010FFBack pressure status 00000000Fusilli Tx Enable 0001FFFFCell count 000BB0E1Reset VOQ XXXXXXXXDMA Busy Status 00000000DMA Done Status 000000FFDescriptor start addr for queue 0 0377F980Descriptor start addr for queue 1 037B5800Descriptor start addr for queue 2 037EB900Descriptor start addr for queue 3 03821980Descriptor start addr for queue 4 03857A00Descriptor start addr for queue 5 0388DA80Descriptor start addr for queue 6 038C3B00Descriptor start addr for queue 7 038F9B80Descriptor start addr for queue 8 0392FC00Descriptor start addr for queue 9 03965C80Descriptor start addr for queue 10 0399BD00Descriptor start addr for queue 11 039D1D80Descriptor start addr for queue 12 03A07E00Descriptor start addr for queue 13 03A3DE80Descriptor start addr for queue 14 03A73F00Descriptor start addr for queue 15 03AA9F80Descriptor start addr for queue 16 03AE0000
The next example shows how to display statistics about the time it takes (in milliseconds) for PSAR to process and send packets to the fabric interface. The statistics in the Timeouts column show how many times PSAR timed out while sending a packet to the fabric. A timeout to a line card may indicate a problem in communication between the PRP and the line card.Router# show controllers psar outputTo Fabric servicing time statistics:----------------------------------Slot Minimum Maximum Average Timeouts0 0 4 0 01 0 4 0 02 0 4 0 23 0 0 0 04 0 4 0 05 0 4 0 06 0 4 0 07 0 4 0 08 0 4 0 09 0 4 0 010 0 4 0 011 0 4 0 012 0 4 0 013 0 4 0 014 0 4 0 015 0 4 0 016 0 4 0 0
CGM—Cisco GSR Manager. An element management system for managing Cisco 12000 Series Routers.
BGP—Border Gateway Protocol. An interdomain routing protocol designed for the global Internet. Exterior border gateway protocols (EBGPs) communicate among different autonomous systems. Interior border gateway protocols (IBGPs) communicate among routers within a single autonomous system.
EBGP—Exterior Border Gateway Protocol. EBGPs communicate among different network domains.
GRP—Gigabit Route Processor.
GSR—Gigabit Switch Router. Also known as Cisco 12000 Series Router.
IGP—Interior Gateway Protocol. Internet protocol used to exchange routing information within an autonomous system. Examples of common Internet IGPs include IGRP, OSPF, and RIP.
MIB—Management Information Base. Database of network management information that is used and maintained by a network management protocol such as SNMP. The value of a MIB object can be changed or retrieved using SNMP commands, usually through a network management system (NMS). MIB objects are organized in a tree structure that includes public (standard) and private (proprietary) branches.
NSF—Non-Stop Forwarding. The ability of a router to continue to forward traffic toward a router which may be recovering from a transient failure. Also, the ability of a router recovering from a transient failure in the control plane to continue correctly forwarding traffic sent to it by a peer.
OSPF—Open Shortest Path First. Link-state, hierarchical IGP routing algorithm proposed as a successor to RIP in the Internet community. OSPF features include least-cost routing, multipath routing, and load balancing.
PRP—Performance Route Processor.
RIP—Routing Internet Protocol. IGP supplied with UNIX BSD systems. The most common IGP in the Internet. RIP uses hop count as a routing metric.
RPR+—Route Processor Redundancy Plus: an enhancement to RPR / EHSA in which the standby RP is fully initialized. An RPR+ switchover does not involve line card reset nor line card software reload.
SNMP—Simple Network Management Protocol. Network management protocol used almost exclusively in TCP/IP networks. SNMP provides a means to monitor and control network devices, and to manage configurations, statistics collection, performance, and security.
SSO—Stateful switchover. SSO provides protection for network edge devices with redundant processors (RPs) that represent a single point of failure in the network design, and where an outage might result in loss of service for customers.