Feature Requirements |
Series |
Models |
|---|---|---|
Before enabling the Add-Path function, you need to set the number of paths for use according to the actual requirement. The maximum number of paths cannot be set; otherwise, memory resources may be exhausted. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
If the AS-SET parameter is configured during route summarization, the AS_Path attributes of all specific routes with the same sequence number form AS_SEQUENCE, and the other ASs form AS_SET, which is used as the AS_Path attribute of the summarized route. The number of AS_Path attributes after aggregation cannot exceed 2000. Otherwise, the AS_Path attribute is set to null. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
LocalIfnet does not support over GRE tunnels. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
In a scenario where a BGP IPv4 VPN unicast route is iterated to an MPLS local IFNET tunnel, the tunnel ID in the FIB table is different from that in the IP routing table. The tunnel ID in the FIB table does not carry a VPN instance ID, and the tunnel ID in the IP routing table is used. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
The peer allow-as-loop command enables BGP to check the count of the local AS number in the routes received from EBGP peers or confederation EBGP peers. The command does not apply to IBGP peers or confederation IBGP peers. If the command is not run, the implementation is equivalent to the peer allow-as-loop 0 configuration. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
EBGP peers do not support the following features: Route reflector, Best-external, Add-path. IBGP peers do not support the following features: EBGP-max-hop, MPLS local IFNET, Fake AS. Feature exclusiveness: The ebgp-max-hop and valid-ttl-hops functions are mutually exclusive. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
To advertise default routes, you need to run both the default-route imported and import-route commands. If either command is not run, default routes cannot be advertised even when they are available in the routing table. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
The undo peer x.x.x.x group and undo peer commands have the same function. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
BGP routes cannot be iterated to SRv6 BE routes. If a BGP route is iterated to an SRv6 BE route, the route becomes inactive. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
(1) Configuration restrictions: A maximum of 100 regions can be configured in one VS. A maximum of 100 AS numbers (either 4-byte or 2-byte) can be configured in one region. A maximum of 100 regional confederations can be configured in one VS. A maximum of 100 regions can be configured in one regional confederation. The same AS cannot be added to different regions. The display format is not affected. For example, 1.0 and 65536 indicate the same AS. The same region cannot be added to different regional confederations. The regions to be added to a regional confederation must exist. If a region has been added to a regional confederation, deleting the region will delete the region's information in the regional confederation. (2) Regional authentication depends on the trustworthiness of the original AS in each route. In actual application, the RPKI ROA function must also be deployed to ensure the correctness of the original AS. (3) The regional authentication configuration is a global configuration. Currently, the scenario where the public and private network ASs overlap is not considered. (4) Regional authentication can be enabled or disabled only in the address family view, not for a single peer. Currently, the following address families support regional authentication: IPv4 unicast address family, IPv6 unicast address family, VPN instance IPv4 address family, VPN instance IPv6 address family, BGP multi-instance VPN instance IPv4 address family, IPv4 unicast label address family, and IPv6 unicast label address family (5) Regional authentication must be configured on a border router that is directly connected to a router in an external region. Risky internal routes that already exist cannot be identified. (6) When both an import policy (for example, to modify AS numbers) and regional authentication are configured, regional authentication is performed before the import policy takes effect (after ROA). (7) Regional authentication affects the route learning performance of EBGP peers (the learning performance cannot decrease by more than 5% after regional authentication is configured). Regional authentication does not affect the convergence performance, RR reflection performance, or IBGP peers' route learning performance. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
The routing policy configured using the apply cost-type med-inherit-aigp command takes effect on the IPv4 and IPv6 private networks. The AIGP attribute of the private network route on the PE is advertised to the CE through the MED attribute. The policy applied to other address families does not take effect. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
Currently, among the 11 types of segments specified in the protocol, only Type 1 SIDs (MPLS label SIDs) and Type 2 SIDs (IPv6 address SIDs) are supported. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
In a public network IP over SRv6 TE Policy scenario, mirror-SID-based egress protection and service FRR-based egress protection are not supported. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
In the SR-MPLS TE Policy address family, IP prefix list, ACL, and RD filters cannot be used to filter routes based on NLRI. If the route-policy configured for a peer in the SR-MPLS TE Policy address family contains the if-match ip-prefix, if-match acl, or if-match rd-filter configuration, all the routes of the peer match the route-policy. If the route-policy configured for a peer in the SR-MPLS TE Policy address family contains the if-match as-path or if-match cost configuration, only the routes that meet the filtering rules match the route-policy. Properly plan route-policies. Do not use unsupported matching modes. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
Load balancing is not supported between MPLS and SRv6 tunnels. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
BGP IPv4 public network unicast routes, BGP IPv4 VPN remote cross routes, BGP IPv4 public network labeled routes (in 6PE networking), and BGP IPv6 VPN remote cross routes can recurse to SR-MPLS TE Policy tunnels based on next hop+color. Other routes do not support such recursion. Properly plan the tunnel policy and tunnel selector. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
By default, after a static route recurses to an SRv6 BE route, the static route becomes inactive. To prevent this problem, run the ip route-static recursive-lookup inherit-label-route segment-routing-ipv6 command. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
BGP IPv4 public network unicast routes, BGP IPv4 VPN remote cross routes, BGP IPv6 public network labeled routes, and BGP IPv6 VPN remote cross routes can recurse to SR-MPLS TE Policy tunnels. If multiple color extended community attributes are carried, only the maximum color value is used for recursion to SR-MPLS TE Policy tunnels. In addition, the value of the CO flag (refer to draft-ietf-idr-segment-routing-te-policy) in each received route is considered 00, regardless of its actual value. Properly plan the color extended community attribute to be added to routes on the egress. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
Allocation of DX4 SIDs for BGP public network IPv4 routes by the next hop is mutually exclusive with the Add-path function. If Add-path has been enabled before a DX4 SID is applied for, no DX4 SID can be obtained. If a DX4 SID has been obtained before the Add-path function is enabled, the allocated DX4 SID is released. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
After a route is imported between VPN and public network instances, the next hop or color extended community attribute of the route cannot be changed through a route-policy. Properly plan the route-policy to be used during the route import between VPN and public network instances. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
IPv4 and IPv6 neighbor functions cannot be enabled at the same time in the VPNv4 address family. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
When the apply extcommunity soo command is run in the XPL policy and route-policy, only the apply extcommunity soo {<source-of-origin> &<1-16>} additive command takes effect. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
Scenario 1: The as-number command is run in the BGP-VPN instance IPv4 or IPv6 address family view to configure a VPN AS number, and the bgp yang-mode enable command is run to enable the YANG management mode for the BGP VPN instance. Restrictions: 1. After the bgp yang-mode enable command is run, the sequence of the peer as-number and as-number commands in the configuration file changes. As a result, configurations in the configuration file fail to be pasted. To solve this problem, configure a VPN AS number using the as-number command first and then run the other commands in the configuration file. 2. Do not configure BGP peer relationships in the BGP-Flow VPN instance IPv4 address family view, BGP-Flow VPN instance IPv6 address family view, or BGP-labeled-VPN instance IPv4 address family view. Do not configure BGP-VPN IPv4 peer relationships in the BGP-VPN instance IPv6 address family view. 3. Before deleting the VPN AS number, delete all the peers and peer groups in the BGP-VPN instance IPv4 or IPv6 address family as well as all the peers in the BGP VPN instance view. 4. Before running the undo bgp yang-mode enable command, delete the VPN AS number. Scenario 2: A VPN AS number is configured in both the BGP-VPN instance IPv4 address family view and BGP-VPN instance IPv6 address family view, and the YANG management mode is enabled for the BGP VPN instance and then disabled. Restrictions: 1. After the bgp yang-mode enable command is run, the sequence of the peer as-number and as-number commands in the configuration file changes. As a result, configurations in the configuration file fail to be pasted. To solve this problem, run the peer as-number command in both the BGP-VPN instance IPv4 address family view and BGP-VPN instance IPv6 address family view to configure a new peer or peer group. However, do not configure a new peer in the BGP-VPN instance view. 2. Do not configure BGP peer relationships in the BGP-Flow VPN instance IPv4 address family view, BGP-Flow VPN instance IPv6 address family view, or BGP-labeled-VPN instance IPv4 address family view. Do not configure BGP-VPN IPv4 peer relationships in the BGP-VPN instance IPv6 address family view. 3. Before deleting the VPN AS number, delete all the peers and peer groups in the BGP-VPN instance IPv4 or IPv6 address family as well as all the peers in the BGP VPN instance view. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
The precautions for BGP egress peer engineering (EPE) are as follows: 1. Enabling EBGP peers is only supported. 2. The confederation member ASN is not supported. 3. The peer group cannot be enabled. 4. The directly connected EBGP peer generates a node label, not an adjacency label. 5. The BGP EPE takes effect only after the BGP-LS address family is enabled. 6. BGP EPE supports only two direct connection scenarios. Multi-hop EBGP supports only two direct connections. 7. The BGP EPE label forwarding function takes effect only after segment routing is enabled in the system view. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
After the peer fake-as command is run and the prepend-fake-as or prepend-global-as parameter is modified, the BGP peer relationship is reestablished. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
If IPv6 routes received from IPv6 peers recurse to 6PE routes, the Pop-Go forwarding mode is not supported. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
Mode in which a tier-2 carrier network accesses a tier-1 carrier network: IGP+LDP; Load balancing or FRR is not supported for traffic from a tier-2 carrier network to a tier-1 carrier network. 1 -> n+1 |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
The original community values in all BGP public network routes on a device must be different from the configured community values that indicate peer roles. Otherwise, the original community attributes will be replaced, or routes will be incorrectly filtered. Users need to ensure that the community values are unique during the overall network planning. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
During BGP peer relationship establishment, specifying the local address is only optional. Therefore, when binding a TWAMP Light test instance to a connection based on the virtual link of the peer, the device cannot check whether the local and remote addresses of the peer match those of the TWAMP Light test instance. However, the generated link routes do not contain delay information. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
Only End/End.X SIDs support compression. Other types of SIDs do not support compression. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |
UNRs do not support POPGO. |
NetEngine 8000 F |
NetEngine 8000 F2A/NetEngine 8000 F1A |