BGP

BGP Policy Accounting

Border Gateway Protocol (BGP) policy accounting measures and classifies IP traffic that is received from different peers. You can identify and account for all traffic by customer and bill accordingly.

Policy accounting is enabled on an individual input interface basis. Using BGP policy accounting, you can now account for traffic according to the route it traverses.

This feature is now supported on routers that have the Cisco NC57 based line cards with external TCAM (eTCAM) and operate in native mode.

This feature introduces these changes:

• CLI: The feature introduces the hw-module fib bgppa stats-mode command.

• YANG Data Model: New XPaths for Cisco-IOS-XR-um-hw-module-profile-cfg.yang (see GitHub, YANG Data Models Navigator)

Detect Slow Peer in a BGP Group

BGP peers process the incoming BGP update messages at different rates. A slow peer is a peer that is processing incoming BGP update messages very slowly over a long period of time compared to other peers in the update sub-group.

Slow peer handling is important when routes are constantly changing over a long period of time. It is important to clean up stale information in the queue and send only latest state. It is helpful to know if there is a slow peer, which indicates there is a network issue, such as sustained network congestion or a receiver not processing updates on time, that the network administrator can address.

Programmability

Stream Telemetry Data for ACL Byte Counters

You can stream model-driven telemetry (MDT) data to monitor the ACL statistics such as stopped, matched and denied IPv4 and IPv6 packets using Cisco-IOS-XR-ipv4-acl-oper.yang and Cisco-IOS-XR-ipv6-acl-oper.yang data models. This release lets you stream telemetry data to monitor the statistics using byte counters.

Previously, the only option to monitor ACL statistics was to use packet counters.

ACL with policer statistics is supported only on Cisco Network Convergence System 5700 Series Routers.

Securely retrieve dynamic NACM with LDAP over TLS authentication

You can now securely retrieve the NETCONF Access Control Model (NACM) policies or rules on-demand from a remote Lightweight Directory Access Protocol (LDAP) server to validate each NETCONF operation using Transport Layer Security (TLS) authentication. With TLS authentication, the router and the LDAP server communication is encrypted for security. Before this release, the policies or rules were not encrypted and posed security vulnerabilities.

Disable TLS Version 1.0

Although Transport Layer Security (TLS) provides secure communication between servers and clients, TLS version 1.0 may pose a security threat. You can now disable TLS version 1.0 using the tlsv1-disable command.

Interface and Hardware Component

Configure lower port speeds for dual-mode optical modules

You can now configure the lower port speed using simple CLI keyword: speed or quad and switch between the higher and lower speeds without changing the optical module.

Earlier, by default, only the higher port speed was available.

The feature introduces new XPaths for YANG Data Model: Cisco-IOS-XR-optics-speed-cfg.yang

(see GitHub, YANG Data Models Navigator.)

Transmission of VLAN-Tagged LLDP Packets

With this release, transmitting VLAN-tagged LLDP packets on the subinterfaces is supported. Earlier, if LLDP is enabled on a subinterface, the LLDP packets are sent without a VLAN tag.

VLAN-tagged LLDP packets help to identify unauthorized devices on the network and discover VLANs configured on the network devices. You can monitor and enforce VLAN segregation, ensuring that devices are connected to the correct VLANs and preventing unauthorized access to sensitive network segments.

You can enable VLAN tagging for LLDP packets globally or on each subinterface using these commands:

• Globally: lldp subinterfaces-tagged

• Each subinterface: lldp tagged

Two-pass Forwarding over BVI

With this release, Integrated Routing and Bridging/Bridge-group Virtual Interface (IRB/BVI) supports Layer 2 ACL, QoS, and statistics on BVI-routed packets, using a two-pass forwarding model for packets over BVI.

This feature introduces the following changes:

• CLI: hw-module irb

• YANG Data Model: New XPaths for modules Cisco-IOS-XR-fia-hw-profile-cfg.yang and Cisco-IOS-XR-um-hw-module-profile-cfg.yang

oFEC Traffic Configuration for QDD-400G-ZRP-S

QDD-400G-ZRP-S optical module can now support the following oFEC traffic configurations:

• 400G-TXP-1x1 DAC-16 QAM

• 3x100G-MXP-1x1 DAC-8 QAM

• 2x100G-MXP-1x1.25 DAC-8 QAM

• 2x100G-MXP-1x1.25 DAC-16 QAM

This increases the interoperability of the QDD-400G-ZRP-S optical module across network components supporting these formats.

IP Addresses and Services

DHCP Snooping for Layer 2 networks

With this feature, you can secure your DHCP infrastructure for Bridge Domains. DHCP Snooping operates in the Layer 2 network and prevents unauthorized DHCP servers from accessing your network.

This feature mitigates the security risks due to denial-of-service from rogue DHCP servers, which disrupt networks as they compete with legitimate DHCP servers that configure hosts on the network for communication.

You can use the following data models to configure this feature:

• Cisco-IOS-XR-ipv4-dhcpd-oper.yang

• Cisco-IOS-XR-l2vpn-oper.yang

• Cisco-IOS-XR-um-dhcp-ipv4-cfg.yang

Display ACL Statistics in Bytes

We have enabled better visibility of traffic distribution, thus helping you in capacity planning, network optimization, and identifying potential bottlenecks in network planning by displaying ACL statistics in bytes in ingress and egress directions. Previously, the statistics were available only in packet counts. The ACL statistics in bytes addition to packet count, help identify the average package size in the network and detect if the packets are truncated or not.

You can view the ACL statistics in bytes for ACL-Based Policing only in Cisco NCS 5700 Series Routers and Cisco NC 57 line cards installed and operate in native and compatibility mode.

The following commands are modified in this feature:

• show access-lists ipv4

• show access-lists ipv6

Limit Address Resolution Protocol (ARP) Cache Entries per Interface

In this feature, you can configure the maximum limit for the number of entries of dynamic mapping between IP addresses and media addresses by ARP per interface. Limiting the number of entries provides overflow protections in ARP cache and protects the routers from DOS attacks by preventing memory overuse by cache entries.

This feature introduces the arp cache-limit command.

Rate Limiting the Multicast and Broadcast Punted Traffic at Subinterface level

When an Ethernet Virtual Connection (EVC) on a port is stormed with multicast or broadcast punted traffic, it impacts the performance of all the other EVCs on that particular port due to the NPU resource sharing. You can avoid such situations using rate limiting at subinterface level for the multicast and broadcast punted traffic.

This feature is supported on routers that have the NC57 SE (Services Edge Optimized) version line cards installed and operating in native mode.

L2VPN and Ethernet Services

Call Admission Control for L2VPN P2P Services over Circuit-Style SR-TE Policies

This feature allows you to configure guaranteed bandwidth for Layer 2 P2P services steered over Circuit-Style SR-TE policies.

This ensures that a Circuit-Style SR-TE policy has sufficient bandwidth to accommodate a Layer 2 P2P service, while also preventing a L2 P2P service from being steered over a Circuit-Style SR-TE policy when there is insufficient available bandwidth.

Dynamic Address Resolution Protocol (ARP) Inspection (DAI)

The routers can now determine the validity of an Address Resolution Protocol (ARP) packet based on valid MAC address to IP address bindings stored in a trusted database built at runtime by DHCP snooping.

With this feature, the router relays only the valid ARP requests and responses, thus preventing the ARP poisoning attacks.

This feature introduces the following:

• CLI: New dynamic-arp-inspection command.

• Yang Data Model: Cisco-IOS-XR-l2vpn-oper.yang and Cisco-IOS-XR-ipv4-arp-oper.yang

IP SourceGuard (IPSG)

You can now achieve source IP address filtering on a Layer 2 port, to prevent a malicious host from manipulating a legitimate host by assuming the legitimate IP address of the host. This feature uses dynamic DHCP snooping and static IP source binding to match IP addresses to hosts.

This filtering limits the ability of a host to attack the network by claiming the IP address of a neighbor host.

This feature introduces the following:

• CLI: New ip-source-guard command.

• Yang Data Model: Cisco-IOS-XR-l2vpn-oper.yang

Set EVPN Gateway IP Address in EVPN Route Type 5 NLRI

You can now set the EVPN gateway IP address in the EVPN route type 5 network layer reachability information (NLRI) that advertises IPv4 and IPv6 addresses. By setting the EVPN gateway IP address, only one IP-MAC route is withdrawn ensuring fast traffic switchover and reduced convergence time in the event of failure. Furthermore, this feature facilitates optimal traffic load balancing across the Virtual Network Forwarders (VNFs) and minimizes control plane updates when the VNFs or virtual machines (VMs) move.

Previously, the NLRI advertisement included the default EVPN gateway IP address of zero, which was represented as 0.0.0.0 for IPv4 and :: for IPv6. This resulted in the withdrawal of all prefixes one by one in the event of a failure, leading to traffic loss.

This feature introduces the following new commands:

• set advertise-evpn-gw-ip

• advertise gateway-ip-disable

L3VPN

Resource Optimization for MPLS Inter-AS Option B

You can now preserve MPLS encapsulation ID resources for Inter-AS option B local labels on Cisco NCS 5700 Series routers with Embedded Ternary Content-Addressable Memory (eTCAM) cards without the need to allocate any additional resources for these IDs. This feature is enabled by default and you cannot disable it.

Previously, these encapsulation IDs were allocated but left unused.

Multicast

Load Balancing in Unicast GRE Tunnels

We now support the Equal-Cost Multipath (ECMP) and Link Aggregation Groups (LAG) load-balancing techniques for transporting multicast traffic over unicast GRE tunnels.

ECMP and LAG provide higher bandwidth and redundancy, better network performance, and fault tolerance by using all available links.

Multicast Traffic Over Layer 2 IPv6 Network

This feature is supported on routers that have the Cisco NC57 line cards installed and operate in native and compatible modes.

Routers use Multicast Listener Discovery (MLD) protocol to discover the devices in a network and create route entries or update the route status in an IPv6 multicast network.

Modular QoS

Additional Routers Supported for ACL with Fragment Match

You can now prevent malicious users from staging denial of service (DoS) attacks for non-initial IP fragments on Cisco NCS 5700 Series Routers and on NCS 5500 Series Routers with line cards other than NC57 line cards.

You can achieve this by configuring an ACL with fragment match and specifying QoS match actions to rate-limit noninitial fragments for IPv4 traffic.

Previously, this functionality was available only on systems with NC57 line cards running in native mode.

Routing

Flexible-Algorithm Redistribution in IP Networks

You can now select or filter the prefix-matching algorithm number during route redistribution, so that only the Flex-Algorithms that you configured for specific addresses are redistributed.

This feature introduces the set algorithm command.

Limiting LSA numbers in a OSPF Link-State Database

The nonself-generated link-state advertisements (LSAs) for a given Open Shortest Path First (OSPF) process is limited to 500000. This protection mechanism prevents routers from receiving many LSAs, preventing CPU failure and memory shortages, and is enabled by default from this release onwards. If you have over 500000 LSAs in your network, configure the max-lsa command with the expected LSA scale before upgrading to this release or later.

This feature modifies the following commands:

• show ospf to display the maximum number of redistributed prefixes.

• show ospf database database-summary detail to display the number of LSA counts per router.

• show ospf database database-summary adv-routerrouter ID to display the router information and the LSAs received from a particular router.

Limiting the Maximum Redistributed Type-3 LSA Prefixes in OSPF

By default, the maximum redistributed Type-3 LSA prefixes for a given OSPF process is now limited to 100000. This mechanism prevents OSPF from redistributing a large number of prefixes as Type-3 LSAs and therefore preventing high CPU utilization and memory shortages.

Once the number of redistributed prefixes is reached or exceeds the threshold value, the system log message is generated, and no more prefixes are redistributed.

Segment Routing

IS-IS: Flexible Algorithm Reverse Affinity

This feature enhances the IS-IS Flexible Algorithm link admin group (affinity) constraint to include link colors on links in the reverse direction toward the calculating router.

The ability to apply affinity constraints in the reverse direction provides additional control for IS-IS Flexible Algorithm path computation.

SR-TE Automated Steering Without Service Label

This feature allows traffic to a BGP service route to be steered over an SR-TE policy using the AS principles, and without imposing the service route’s prefix label.

This feature enables use-cases such as centralized BGP EPE for 6PE in an SR-MPLS network.

This feature introduces the following command:

• bgp prefix-label ignore

SR-TE Explicit Segment Lists with Mix of IPv4 and IPv6 Segments

Explicit segment list can be configured to include IPv6 segments, for example IPv6 adjacency SIDs or IPv6 EPE SIDs.

This feature enables use-cases such as Centralized BGP EPE for 6PE in an SR-MPLS Network.

SRv6 Services: Services with Remote SIDs from W-LIB

This feature enables an SRv6 headend node to receive and install remote SIDs with Wide (32-bit) functions (Remote W-LIB).

There is no new CLI to enable this capability at the ingress PE.

System Management

Auto-Save with Secure File-Transfer and Additional Configurable Parameters

Apart from automatically backing up the running configuration after every commit, you can also do the following with Auto-Save:

• Save running configurations to remote systems using Secure Copy Protocol (SCP) and Secure File Transfer Protocol (SFTP).

• Configure wait-time between two subsequent auto-saves.

• Append time-stamp to the file name of the saved configuration.

• Save the encrypted password.

• Specify the maximum number of files that you can auto-save.

The feature introduces these changes:

CLI: Modified the configuration commit auto-save command by adding the following keywords:

• filename scp

• filename sftp

• wait-time

• timestamp

• password

• maximum

Yang Data Model:

• New XPaths for Cisco-IOS-XR-config-autosave-cfg

• New XPaths for Cisco-IOS-XR-um-config-commit-cfg

FQDN for NTP Server on Non-default VRF

You can now specify a Fully Qualified Domain Name (FQDN) as the hostname for NTP server configuration over non-default VRFs.

FQDNs are easy to remember compared to numeric IP addresses. Service migration from one host to another can cause a change in IP address leading to outages.

Prior releases allowed FQDN handling in only default VRFs.

Support for SFTP (Secure File Transfer Protocol) and SCP (Secure Copy Protocol) options in the Copy command

With this feature, the router can transfer data to a remote server on SFTP and SCP by using the underlying SSH protocol implementation. You can use the SFTP and SCP option to facilitate secure transfer of configuration files from the router to an achieve server.

This feature modifies the copy command.

GNSS MIBs Traps for Antenna Open-Circuit, Satellite Visibility and Module Lock Status

Your router uses Global Navigation Satellite System (GNSS) as the satellite system for enhanced timing synchronization of the timing interface. GNSS receiver picks up signals from this satellite system to recalculate position, velocity and local time to high precision.

From this release, you can track the GNSS module antenna OC alarm status, status of the GNSS satellite visibility, and specify the lock status of GNSS module.

You can useMIB Navigator tool to know more about the following traps introduced in this release:

• ciscoGnssAntennaOCAlarmStatus

• ciscoGnssSatelliteVisibilityStatus

• ciscoGnssModuleLockStatus

Enhanced SyncE and extended ESMC

ITU-T G.8262.1 recommendation defines the requirements for timing devices used in synchronizing network equipment. For example, bandwidth, frequency accuracy, holdover, and noise generation.

With Enhanced SyncE (eSyncE) and Extended Ethernet Synchronization Message Channel (eESMC) support, the routers are capable of handling the following SyncE clocks on the network:

• Enhanced ethernet equipment clock (eEEC)

• Enhanced primary reference clock (ePRC)

• Enhanced primary reference timing clock (ePRTC)

This feature is supported on the following NCS5500 and NCS 5700 variants:

• Cisco NCS55A2

• Cisco NCS 57C3

• Cisco NCS-57B1

• Cisco NCS-57C1

New Cisco-NTP-MIB Traps to Monitor NTP server and Improve Timing Accuracy

Cisco-NTP-MIB allows you to monitor NTP on the server and client using SNMP MIB. This release supports new traps, which will help monitor the NTP server and improve timing accuracy . These traps also display the NTP server's current status, the local clock's stratum, the maximum error in seconds, and the delay in round-trip in seconds.Use MIB Navigator to know more about the newly added traps:

• cntpSysPeer

• cntpSysSrvStatus

• cntpSysStratum

• cntpSysRootDelay

• cntpSysRootDispersion

• cntpPeers

System Security

Accessing Certificate Enrollment URL Using HTTP Proxy via specified Source Interface

With this feature, you can enable the router to use an HTTP proxy to access the certificate enrollment URL. The router uses the already available HTTP proxy configurations to fetch Certificate Revocation List (CRL) to access the certificate enrollment URL. In addition, you can specify a source interface through which the router places the enrollment requests.

This feature reduces the enrollment URL access failures when the router fails to reach the enrollment URL directly or when the enrollment URL is only reachable via an HTTP proxy.

IEEE 802.1X Port-Based Authentication Support for Multiple Authentication and Multiple Host Modes

The IEEE 802.1X port-based authentication allows only authorized supplicants to access the network. The IEEE 802.1X port-based authentication now supports multiple authentication and multiple host modes to allow multiple hosts or MAC addresses on a single port.

Applicable to the following Cisco NCS 5500 Series Routers:

• NCS-57C3-MOD-SYS

• NC57-36H6D-S

• NC57-MOD-S

Applicable to the following Cisco NCS 5700 Series Router:

• NCS-57C1-48Q6-SYS

Secure Key Integration Protocol (SKIP) for Routers

The NCS 5500 Series Routers are now capable of handling the SKIP protocol. With this ability, it can communicate with external quantum devices. This helps in using Quantum Key Distribution (QKD) devices for exchanging MACsec encryption keys between routers to eliminate the key distribution problem in a post quantum world where the current cryptographic systems are no longer secure due to the advent of quantum computers.

This feature introduces the following:

• CLI:

  • crypto-sks-kme

  • show crypto sks profile

  • show crypto sks profile stats

• Yang Data Model: Cisco-IOS-XR-um-sks-server-cfg.yang

For more information on Quantum Key Distribution, see Post Quantum Security Brief.

Securely retrieve NACM policies using LDAP over TLS connection

You can now securely retrieve the NETCONF Access Control Model (NACM) policies or rules from a remote Lightweight DirectoryAccess Protocol (LDAP) server using Transport Layer Security (TLS) authentication. With TLS authentication, the communication between the router and the LDAP server is encrypted for security.

Before this release, the communication between the LDAP server and the router was not secured.

Programmability

Cisco-IOS-XR-config-autosave-cfg

This Cisco native YANG data model enables you to automatically backup the running configuration files after every commit is made.

Cisco-IOS-XR-um-config-commit-cfg

This unified data model enables you to automatically back up the running configuration of the routerafter every commit is made.

openconfig-network-instance.yang Version 0.2.3

In this release, the installed counter in the OpenConfig data model is enhanced to view the number of routes that are installed in Routing Information Base (RIB) from a specific neighbor per Address Family Identifiers (AFI) or Subsequent Address Family Identifiers (SAFI).

The model supports a single instance of BGP with default VRF, and IPv4/IPv6 address family. Cross AFI where an IPv4 route learnt from an IPv6 neighbor, and vice versa, is not supported.

You can stream Event-driven telemetry (EDT) and Model-driven telemetry (MDT) data.

Cisco-IOS-XR-um-mpls-static-cfg.yang

This Unified data model enables you to configure Label Switched Paths (LSPs) with statically assigned ingress labels that are mapped to prefixes or VRFs and egress paths explicitly defined or mapped to next hops. With this release, you can use the data model to disable the default route to resolve issues with the next hop information.

We recommend that you use the data model according to the CLI hierarchy.

Cisco-IOS-XR-remote-attestation -agent-oper.yang

This Cisco native data model defines the remote attestation of the routers' security posture to assess the trustworthiness of hardware and software on the router.

With this release, you can use the data model to send gNMI requests to retrieve the system integrity information such as secure boot status,Attestation Identity Key (AIK) and Secure Unique Device Identifier (SUDI) certificates.

gNMI support is introduced for the show platform security attest certificate command.

openconfig-isis.yang Version 1.0.0

The OpenConfig data model is revised from version 0.6.0 to 1.0.0 to simplify the authentication keychain nodes. With this feature, you can configure the authentication type to limit the establishment of adjacencies and the exchange of LSPs. You can also retrieve the operational state of the authentication nodes.

openconfig-network-instance.yang

In this release, the installed counter in the OpenConfig data model is enhanced to view the number of routes that are installed in Routing Information Base (RIB) from a specific neighbor per Address Family Identifiers (AFI) or Subsequent Address Family Identifiers (SAFI).

The model supports a single instance of BGP with default VRF, and IPv4/IPv6 address family.

You can stream Event-driven telemetry (EDT) and Model-driven telemetry (MDT) data.

openconfig-bgp.yang Version 9.1.0

This OpenConfig data model, which is part of the openconfig-network-instance.yang data model is revised from version 6.0.0 to 9.1.0. This version introduces the following changes:

• Use the restart-time node to configure the time interval (in seconds) to reestablish a disconnected BGP session when the maximum prefix limit is exceeded.

• Remove the restart-timer node that is used to set the maximum restart time for local BGP sessions from global and neighbor Address Family Indicator (AFI) and Subsequent Address Family Indicator (SAFI) sessions.

Cisco-IOS-XR-optics-speed-cfg.yang

This Cisco native YANG data model is introduced to configure lower port speeds for dual-mode optical modules.