ccnp tshoot hsrp sim

ccnp tshoot hsrp sim

The maximum time allowed before a KEEPLIVE or update message must be received . If Holdtime is inconsistent at both ends , both parties accept a smaller time. After the Uplinkfast feature is activated , the switch automatically adjusts some parameters: Static route associated with the outbound interface GLBP We look at the above picture, the external detail route 172.16.32.0 – 39.0/24 , and the summary route 32.0/21 are introduced into OSPF by R1 , now Match the previous character 0 or more times STP and RSTP port comparison: 3.3.3.3 10.1.23.3 from 10.1.23.3 (3.3.3.3) 3.3.3.3 Interface fast0/0 : Disabled Configure all interfaces as passive and manually activate specific interfaces: AS_PATH , and AS_PATH only when routing out this AS or BGP routers to update EBGP will change when neighbors in AS inside will not change, therefore, in the AS internal, anti-ring requires the use of split horizon. So how to solve the problem that R5 ca n't learn routing? IP routing table ( IP Routing the Table ) Let's analyze that SW1 inserts the DHCP request message from the PC into option82 and then sends it to SW2 . Via SW2 Match ip address 2 set metric 3 Routing strategy needs (reliability, redundancy, offloading model, etc.) Is 6 , then the type is 802 , default ethernet After completing the configuration, R2 show ip route Configuration example 1 ( -RIP in a single routing protocol environment ) Since the flash space is very limited, it is recommended to store the file on the TFTP server. When deploying on a TFTP server, be careful to first create an empty file to correspond to the URL in the configuration command . This file is used for dhcp snooping bingding database writes (depending on the device). ACL Bgp dampening half-life reuse limit limit limit maximum inhibition time 10.1.25.2 BPDU format and operation R4 released the RIP route of 100.1.1.0/24 , which can be learned by both ASBRs . The solution is to configure static summary routes on the two ASBRs : Origin IGP, metric 0, localpref 100, valid, internal, best Experimental example: Called when republishing 3.3.3.3 ( metric 2 ) from 3.3.3.3 (3.3.3.3) A router that does not support the route reflector function can act as a single router cluster or as a client The configured commands are simple: So first on R1 and R2 , in order to allow them to reach the server 10.1.1.0 and 2.0 network segments, you need to configure two static routes: Hard reset Ip dhcp snooping vlan 10 Interface fast0/1 Distance 130 2.2.2.2 0.0.0.0 1 / / View the routing flip situation Ip prefix-list 1 seq 5 permit 1.1.1.0/24 route-map test permit 1 Switchport private-vlan host-association 100 102 Standby is 10.1.1.252, priority 100 (expires in 7.572 sec) Bgp confederation identifier 345 bgp confederation peers 64512 neighbor 4.4.4.4 remote-as 64512 Metric Passive-interface Fa0/0 neighbour xxxxxx (global configuration command) activates portfast features on all non- trunk interfaces 10.1.34.3 from 10.1.34.3 (3.3.3.3) Switch(config-if)# ip address 192.168.10.254 255.255.255.0 The weight of the BGP route generated by the local summary is also 32768. !! This is the direct link between R1-R3 The process of processing a VLAN frame by an access port is as follows: On: channel member without negotiation ( no protocol ) ( 0xFFFF0000-0xFFFFFFFF ) is a reserved community value. In addition to these values, several well-known values ​​such as no-export , local-as, etc. are defined. For specific values, see RFC . Router bgp 123 Switch(config)# vlan 255 Switch(config)# interface fast0/1 R3#show ip bgp It takes effect on routes re-advertised from other dynamic routing domains. It does not take effect on direct routes (still 0 ) General BGP routers do not accept containing its own AS routing information number, but can neighbor routing sources neighbors allowas-in a few of my as numbers , to make an exception. Below, let's look at an example: Flex links and Port-Security are not compatible with each other Unreliable based on broadcast Generally seen between switches or between switches and routers and servers It is allowed to write multiple next hop IPs , but these IPs must be the interface IP of the directly connected router . Configuration last modified by 0.0.0.0 at 3-5-93 02:01:49 Send TCP packet *> 100.0.2.0/24 Router bgp 64512 Set the interface to DTP dynamic negotiation, optional auto or desirable Decision summary byte ICMP redirect 8 The configured static network path is redistributed into OSPF . In this way, when the intranet user needs to access the Netcom resources, because of the details of the network access, go to OR1 . To access other resources, take the default route to OR2 . The configuration of R1 is as follows: Avoiding ARP and its vulnerabilities through specific mechanisms in IPv6 Send Keepalive bag [123].[7-9] Statically configure a dhcp snooping bingding dabase entry. The command is configured in privileged mode. By modifying LOCAL_PREF to affect the flow properties The KEEPALIVE message contains only 19 bytes of BGP headers, and does not contain any other data. technical background NEXT_HOP on shared Media (operation on shared media) version R3 associates the route-map with the redistribution of the direct route 3.3.3.3/32 , tag#11 , and the route is passed to R2 . On R2 , the redistribution of RIP to OSPF is deployed. After the external route of 3.3.3.3 is injected into OSPF , the tag value is carried by default. Match community 11 In CISCO IOS , CEF can hash the source and destination IP addresses and import the hashed results into a load balancing table for load balancing. This table has 16 hash buckets , each of which points to an adjacency, and multiple buckets can point to an adjacency at the same time. You can use the show ip cef route prefix internal hidden command to see: Configure an interface as a passive interface: ---- -------------------- BGP table on R2 : 10.1.12.2 The configuration of R3 is as follows: The RRs have physical connections and establish BGP peer relationships. In the case where one of the RRs fails, the client still has an alternate connection. Switch(config-if)# switchport mode encapsulation {dot1q | ISL} Ip dhcp snooping Not advertised to any peer Initially, R3 can learn three loopback routes from R1 and 192.168.12.0/24 routes. Now we don't want R3 to learn the route to 192.168.3.0/24 , then we can configure it on R2 as follows: VTP version running VTP Domain Name VTP Pruning Mode VTP Traps Generation 100 i 0 00:11:53 If the ibgp keyword is not associated , only EBGP routes will be equivalently load balanced (by default only for EBGP routes). If maximum-paths is not configured , the next routing principle will be followed. V1 defaults to send v1 to receive v1, v2 In doing polymerization route, using as-set produced after AS-Path list {} in the AS length number only count a AS length number within the Union AS-Path list () in the AS number length Calculate basis Ip address 10.1.13.1 255.255.255.0 And SW3 on the same case Network 100.0.2.0 mask 255.255.255.0 Match ip address 2 No synchronization no auto-summary Switch(config)# ip routing Ip community-list 11 permit 100:11 The "original Ethernet frame" is up to 1518 bytes, the IP MTU of 1500 , plus the source destination MAC address is 12 , the type field is 2 , CRC4 , plus 30 bytes of ISL encapsulation, which is 1548 bytes. Weight Route-map WT2 permit 20 match ip address prefix-list 2 set weight 200 R3 LocPrf 3.3.3.3 0xFFFF After the basic BGP configuration is complete , R1 establishes an EBGP neighbor relationship with R3 , R2, and R5 . R3 establishes an IBGP neighbor relationship with R4 , R4, and R5 (using the LOOPBACK interface as the update source). Use decimal. RFC format is 12 : 10000 , hexadecimal is 0x 000C2710 , and then converted to decimal 796432 R2(config)# access-list 1 deny 192.168.3.0 R2(config)# access-list 1 permit any R2(config)# router ospf 1 The "original Ethernet frame" is up to 1518 bytes, the IP MTU of 1500 , plus the source destination MAC address is 12 , the type field is 2 , CRC4 , plus 30 bytes of ISL encapsulation, which is 1548 bytes. Experiment 1 Router bgp 200 Use clear port-security dynamic command to clear all port-security interface by dynamically learned secure address entry using clear port-security sticky command to clear all sticky secure address entries Gateway of last resort is 172.16.12.1 to network 192.168.1.0 Run RIP between R1 , R2 , and R3 , and R1 and R2 are injected into their respective loopback ports. When the interface is up/down , all dynamically learned MAC address entries will be cleared. The statically configured secure address entries remain. Finally, look at DP , which is also relatively simple as shown. We take SW3 and SW4 between the segment , for example, due SW3 and SW4 will have BPDU , and SW3 issued SW4 of BPDU obviously superior to SW4 this interface to send generated from its own BPDU , so the final SW4 interfaces win. CDMA-Ix CDMA Ix interface Aggregate-address summary address summary-only as-set When a data frame is sent from the Hybrid port, the switch determines whether the attribute of the VLAN on the port is Untag or Tag . If it is Untag , first strip the VLAN tag of the frame and then send it ; if it is a Tag , send the frame directly. 100 Topology change process Router(config)# router ospf 1 TCAM table Use set excommunity cost x to set in route-map The RID is the largest IP address on the router and tends to be assigned to the loopback address. It can also be set manually by the bgp router-id command. If a path contains an RR attribute, the route generator ID ( originator ) will replace the RID in the optimal path selection process. Network Next Hop Metric LocPrf Weight Path / / Map the port to pVLAN Decision making Match community 1 If the static route configured by the outbound interface is configured, the router considers the destination network segment as a local direct connection . Therefore , when the RIP is in the network , it will be announced. IP routing table ( IP Routing the Table ) The configuration of R2 is as follows: Bgp client-to-client reflection The number of messages sent by MsgSent to the neighbor The packet is captured as follows ( R3 sends the BGP update package to R4 ): Ip address 10.1.34.3 255.255.255.0 Metric =0 Neighbor 3.3.3.3 update-source Loopback0 neighbor 5.5.5.5 remote-as 64513 We restore the experimental environment to the basic configuration: BGP neighbor relationships are as follows ? RIP Version 2 Protocol Analysis If the static keyword is used , it means that the ARP ACLs enable the implicit deny any mechanism. That is, if ARP packets and ARP ACLs are used. Planning Principles No-advertise ? SW1(config-if)# switchport native vlan 1 *Mar 16 17:09:21.858: DHCPD: inconsistent relay information. Switchport private-vlan host-association 100 102 Cc00.043c.0000 (10.1.1.253) local Switchport private-vlan host-association 100 102 Only vlan1 and vlan10 ? Ip sla monitor 1 These three representative routing re-release models. Port-security configuration basic concept Distance 100 192.168.12.1 0.0.0.0 1 IGRP 300, ( aggregated by 300 3.3.3.3 ) // a summary route generated from AS300 , a BGP router 3.3.3.3 produced ? 32 bits Same interface mode ( access , trunk ) 10.1.23.2 Ftp: Database agent URL 100 i Routing VS. Switched Campus Network Architecture Set ip next-hop 11.1.1.2 exit Clear the dynamically learned security address entries on the interface. Shows all routes received from the specified neighbor (a subset of the previous command, shown here is the route remaining after the inbound policy is executed) In this step, we run the BGP protocol on the above basis to complete the establishment of basic BGP neighbor relationships: The experiment was tested in the IOS C3640-IK9O3S-M Version 12.4(25) environment. Weight If you want to turn off ICMP redirects, you need to use no ip redirects on the interface . Frame format R2(config)# router ospf 1 No passive-interface fast 0/24 200 i R1 and R3 , R2, and R5 establish an EBGP neighbor relationship. – LoopGuard and RootGuard cannot be enabled at the same time The other side has received the Keepalive packet, the two sides agreed capability characteristics, started Update on Neighbor 10.1.23.3 dmzlink-bw VTP Operating Mode : Server Maximum VLANs supported locally 1005 Add redundant routing 9 SW1(config-if)# switchport trunk allowed vlan ? Optional delivery to simplify the execution of routing policies Network 3.3.3.3 0.0.0.0 area 0 Route summary issues to be aware of basic concept 2 : Reponse message, send all or part of its routing information, in a Response message R3 configuration is as follows Router ospf 1 Statically configure a dhcp snooping bingding dabase entry. The command is configured in privileged mode. It is usually a direct connection because it will look for direct routes. EBGP default TTL is 1 Switchport trunk encapsulation ? switchport mode trukn switchport nonegotiate Knowledge review ICMP redirect Routing entry for 8.8.8.0/24 \ RIP-2 RIPv2 uses multicast update Network consolidation (same protocol or different protocols) BGP of AS plan shown in FIG. Route 2 : the AS_PATH : 500 the MED : 100 Internal RouterID : 172.16.8.1 Configuration Example Basic knowledge point This is the ICMP redirect message sent by R1 to R3 . Note that the Gateway address field in the header is filled with the next hop IP that is closer to the destination than the destination . VTP operation Policy-based routing is more powerful and flexible than traditional routing. It allows network managers to select forwarding paths based on destination addresses and packet types, packet sizes, applications, or IP source addresses. If the configuration revision number of the VTP client is higher than that of the server , the client can also overwrite the vlan information of the server . R1 and R3 , R2, and R5 establish an EBGP neighbor relationship. Ip dhcp snooping verify mac-address About route reflection clusters Summary 1) 2) two points is MED in IBGP passed without problems (not lost) between, but EBGP between transfer depends on whether the route originated in themselves. ? Neighbor xxxx soft-reconfiguration inbound Route 1 : AS_PATH : 500 When the router is adjacent to the host, they learn in some way. SW1(config)# interface vlan 10 The PC belongs to VLAN 10 and the gateway is on SW2 . The DHCP server belongs to VLAN 100 and the gateway is on SW2 . Native Vlan Type echo protocol ipIcmpEcho 10.2.2.2 source-ipaddr 10.2.2.1 frequency 10 Rcp: Database agent URL Further, the summary is transmitted to the route D after, D will continue to B to pass, and as summarized route from C , thus reaching D of AS400 If an OSPF in- process route cannot be summarized: Match ip address prefix-list 11 set as-path prepend 600 It means for VLAN100 traffic, not to log the ARP ACL discarded packets match Network Both R1 and R2 inject 100.0.1.0 and 100.0.2.0 into BGP using network . Next we SW1 of the Fa0 / 20 interface, do switchport allowed vlan the Remove Trunk 10 , the vlan10 traffic pruned, this time the PC between can not ping pass, go R1 speaking about: Adding redundant routing Redistribute local direct connection network segment into BGP , MED value is 0 ; redistribute local static route into BGP , MED value is 0 Router bgp 100 ? 100 i 8 8 8 8 The IEEE802.3 frame can be divided into the following categories according to the values ​​of the DSAP and SSAP fields: We restore the experimental environment to the basic configuration: BGP neighbor relationships are as follows 200 PROTOCOL Redirection enabled, 597.032 sec remaining (maximum 600 sec) Time to live: 14397.032 sec (maximum 14400 sec) 0xFFFF Use the err-disable command to modify the duration of the err-diasable state. The default is 300S. COST community : Etherchannel port does not support pvlan Bgp client-to-client reflection The configuration of R2 is as follows: ARP spoofing principle Distance bgp xyz //x is the EBGP route, y is the IBGP route, and z is the BGP route announced by the local network . 100 Set as-path prepend 6666 Local-as VTP pruning Set ip address prefix-list RFC1403 BGP OSPF Interaction Default route 108 OSPF routes take effect, and only 1.1.1.0 routes survive. The source of continuous routing is direct routing, or RIP . VTP //trunk package protocol is dot1q R2 , R3 , aiming RIP create static routing domain summary route, point null0 , while only these static routes redistributed into OSPF , and indeed can play a sub-optimal paths to avoid the question, but: ...... Dhcp-snooping LLC sublayer ? Logical aggregation of similar links RIP and IGRP the passive-interface does not send routing updates, but to accept routing updates The switches at both ends of the trunk need to use the same trunk protocol. Method two: neighbor xxx default-originate ? The description of several fields ( ISL header) in the above figure is as follows: The GW configuration is as follows: R3 has two BGP neighbors, one is 10.1.13.1 , which is an EBP peer , and the other is 4.4.4.4 is an IBGP peer . Metric LocPrf Weight s> 172.16.11.0/24 The secure address learned dynamically through voice vlan will not be converted to sticky mac address. Router ospf 1 * i 100 Network 100.0.2.0 mask 255.255.255.0 neighbor 10.1.13.3 route-map test out Router bgp 300 When the switch receives suboptimal BPDUs from other designated switches or root bridges , 802.1D encounters this situation by first ignoring these suboptimal BPDUs . 3.3.3.3 Can be detected when a violation time occurs, and can be punished Identify the type of message: Ip policy route-map PBR 100 i Set ip next-hop verify-availability 1 network entries using 117 bytes of memory 1 path entries using 52 bytes of memory 100 If you turn on always-compare-med Metric=BW+DLY , where DLY is the accumulation of DLY along the inbound interface of the route , and BW is the minimum interface bandwidth. Obviously through DLY Process Switching ( process swap ) 0 00:11:53 IBGP non-equivalent load balancing The LSA sends the neighbor to the neighbor, so the local route is filtered, and the neighbor has (because the neighbor has received the LSA ). *> 100.0.1.0/24 The Layer 2 switch searches for VLAN configuration information based on the destination MAC address and VLAN ID of the frame, and determines from which port the frame is sent out. Compare by default from top to bottom Y Federal configuration and implementation Switch, the switch also forwards it with an updated message to all attached LANs for which it is the designated switch. Site-level site identification: Identify the various sites on the network at the data link layer. Knowledge review ? 100 At the same time, the same OSPF route is learned from both processes and is used on a first come, first served basis. Therefore, if R3 first learns the route from R1 ,ccnp tshoot hsrp sim, it will naturally ignore the one that R2 re-releases. However, if the neighbor relationship between R1 and R3 is already obtained in R3 If there is no tag , mark the PVID of the trunk port and proceed to the next step. Version Path Received WAS, IT A SENDS that the LAN containing the BPDU DATE The up-to-Port Information Stored for that. the In the this Way, the inferior IS Discarded Information, and Superior Information The IS ON propagated Network. " Match community 11 10.1.13.3 VID *>i100.0.2.0/24 Set community no-export additive version number * 192.168.34.3, from 192.168.3.1, 00:09:32 ago, via Serial0/1 Route metric is 20, traffic share count is 1 Dhcp-snooping Switch(config)# interface fast0/15 8 bits BVI Bridge-Group Virtual Interface Internal EIGRP Spanning-tree Encapsulation Status Network 192.168.20.0 0.0.0.255 area 1 So they will be put into the IP global routing table, labeled B . 100 BPDU operation If you do the following configuration on R2 : R1(config-router)#distribute-list 1 in ? // All interfaces As a result, R6 will receive summary routes from R4 and R5 , respectively : 172.16.0.0/16 Switch(config-if)# switchport port-security aging static 8.3.4 Questions 104 Ip address 4.4.4.4 255.255.255.0 Redistribute local direct connection network segment into BGP , MED value is 0 ; redistribute local static route into BGP , MED value is 0 1,10 Optional non-delivery Used to limit the time that the route staying in the routing table has not been updated. IBGP session principle 100 The interface will only receive updates, no updates will be sent. The configuration of R2 is as follows: Protected ports are not able to communicate with each other although they are in the same VLAN . 100 i BGP router identifier 10.1.13.3, local AS number 345 1-4094 Access-list 1 permit 172.16.1.0 route-map unsupp permit 10 match ip address 11 In B and C on a pair of A do strategy, i.e. AS external boundary of AS do MED , such that A selected lower the MED value. ? Interface fast0/0 For example , the configuration of R3 is as follows: Using route reflector 23 Through the above configuration, the BGP neighbor relationship can be established, let's show it: Knowledge review The routing protocols that support TAG are: RIPv2 , EIGRP , OSPF , ISIS , BGP Network Can affect BGP routing (before the equivalent load balancing), somewhat similar to " tie breaker " in the path selection, is to break the tangles when tangled Configuring a sample RIP environment BGP table on R4 : Limit the host connected to the switch port ( filtered according to IP or MAC address) //The default community is not sent, so this command must be configured Set default interface specifies how to send such a packet 3.3.3.3 R1 and R2 run IBGP , use LOOPBACK as the update source, and refer to neighbors . R2 and R3 are EBGP relationships. R3 announces 3.3.3.0/24 into BGP , and configures next-hop-self for R1 on R2 . The format of the resulting file is as follows: S0 DCE Note that the relay device needs to enable the DHCP service: service dhcp SA sends the original switch MAC of the frame R3 has two BGP neighbors, one is 10.1.13.1 , which is an EBP peer , and the other is 4.4.4.4 is an IBGP peer . The process learns the external route, so the update source here is R2 . Path Think carefully about what will be wrong, in fact access-list 1 deny 172.16.32.0 is a summary route 172.16. 32.0/21 0xFFFF Origin IGP, metric 0, localpref 100, valid, external, atomic-aggregate , best Route-map HELLO time Switching is the process of mapping layer 2 to layer 3 addresses andforwarding to a destination interface. Bgp default local-preference 500 // Modify the default lp value of the route originating locally Switch(config-if)# switchport trunk native vlan ? For the same BGP connection, the passwords must be the same; different neighbors can set different passwords. After the authentication is configured, all data segments transmitted through the TCP connection between the peers are verified. In the above figure, we inject OSPF routes into RIP . Traditionally, you can only set metrics on all incoming routes . However, with route-map , we can associate a defined one when configuring re-issue commands. A good route-map , in the route-map , we can create different properties or actions for different routes by creating multiple serial number statements. If deterministic-med is turned on, the information of the route prefix is regrouped by AS. ... (maximum 25 ) R1(config)# ip prefix-list 11 seq 5 deny 11.11.11.0/24 R1(config)# ip prefix-list 11 seq 10 permit 0.0.0.0/0 le 32 R1(config)# router bgp 100 Suboption type Neighbor 10.1.12.2 advertise-map RP1 non-exist-map RP2 Through the above configuration, the BGP neighbor relationship can be established, let's show it: Bgp bestpath as−path ignore This command, if configured, skips this rule Command summary 81 The frame to which the Native VLAN belongs is not tagged when it passes the trunk . Provide a better next hop address. If it is 0.0.0.0 , it means the way to publish this route. Full-duplex MAC : The physical layer operating mode provides access when full-duplex. Ip dhcp snooping binding mac-address vlan vlan-id ip-address interface interface-id expiry seconds If R2 on ip route 1.1.1.0 255.255.255.0 null0 , this time there is no routing table OSPF of 1.1.1.0 routing the unsuccessful thus re-release, so R3 can not learn the 1.1.1.0/24 route; The link has failed. Can also be understood as the maximum life of this BPDU , the default 20S ? //The effect is the same as above 200 300 Metric For link state routing protocols such as OSPF Configuration However, we found that R4 did not pass these two routes to R5 , which is because of the effect of IBGP 's horizontal splitting principle. According to the IBGP split horizon principle, a BGP router, if it learns BGP routes from its IBGP neighbors , will not be able to pass these BGP routes to other IBGP neighbors. The reason for setting this rule is that BGP anti-ring needs to resort to Wherein SW1 is as follows: The switch receives a DHCP packet, the discovery packet of option82 information precisely, the switch from locally originated option82 after information is removed, the DHCP packets forwarded to the DHCP Client . After the policy to take effect View R4 's BGP table: Protocol vulnerability TblVer InQ OutQ Up/Down State/PfxRcd 802.1q No-export 8 bits Experiment 1: host-dependent If there is an EBGP connection between R2 , then a routing loop occurs, and R1 and R2 will receive the summary route and accept the routing update. This is because this summary route does not carry any information about AS100 and AS200 . Atom character example: Often used with the portfast feature on the interface for connecting to the host Switch(config)# spanning-tree vlan-id Configuring static routes Ip address 192.168.10.254 255.255.255.0 Send Update package Ip policy-list as100 permit match as-path 1 NEXT_HOP to 10.1.123.2 , then if C is not on the router to the IP of PVC , will be a problem, so this should be taken into account. Detailed rules Match community 11 Weight R3 summarizes the details of R1 and R2 (summary command plus AS_SET keyword), so that the BGP summary route generated by R3 inherits the detailed AS_PATH attribute, and the packet capture on R4 will find the summary route. = 300 {100,200} AS_PATH : look at the results of the capture, found article summary route AS_PATH attribute consists of two parts ( segment ), the value 300 that the segment is Router# sh ip bgp summary Cluster ID . If the cluster contains multiple RRs , you need to manually configure the cluster ID for each RR . <0-4294967295> 32-bit tag value OSPF BGP table on R4 : When ISP1 is down, and the GW does not detect it (that is , the interface of GW connected to ISP1 is not DOWN ), the traffic of PC access 100 is still thrown to ISP1 , which is disconnected. Second column > Optimal path selected by BGP algorithm STP topology change BGP neighbor We restore the experimental environment to the basic configuration: BGP neighbor relationships are as follows Vlan 100 In AS inside, due to the presence IBGP split horizon, making BGProuter had twenty-two established between IBGP connection, in order to get the full BGP routing updates, but this is a very low scalability approach, but also brought to the network device The two effective ways to solve the IBGP scalability problem are route reflectors and federation. Compared with the federation, the route reflector has the advantage that all routers in the federation need to support and understand the federated mechanism. The route reflector only needs the RR to understand the reflector mechanism. In addition, the implementation mechanism of the route reflector is relatively simple. Of course, if you want to manage large-scale AS with various EBGP mechanisms , then the federation will be a better solution.