Huawei H12-891_V1.0 - HCIE-Datacom V1.0

Understanding BGP EVPN and VXLAN Routing

EVPN (Ethernet VPN) with VXLANis used fordata center networkingandL3 VPN interconnection.

🔹BGP EVPN Route Types & Their Purpose

Type 2:MAC/IP Advertisement Route (Carries MAC + IP)

Type 3:Inclusive Multicast Route (Carries VTEP Multicast Information)

Type 5:IP Prefix Route(Used for Layer 3 VXLAN Routing)

🔹What does Type 5 route contain?

✅Route Distinguisher (RD):Identifies a unique VPN instance.

✅GW IP Address:Specifies theGateway IPfor L3 VXLAN routing.

✅IP Prefix Length:Defines theIPv4/IPv6 network maskfor routing.

❌MAC Address is NOT included in Type 5

Why is Option B (MAC Address) Correct?

✅MAC addresses are carried in EVPN Type 2 routes, NOT in Type 5.

✅Type 5 routes focus on IP prefixes, not MAC forwarding.

🔹Real-World Application:

VXLAN L3 Gateways (EVPN Type 5)allow routers to forwardIP traffic across VXLAN tunnels, removing the need for traditionalL3 VPNs.

✅Reference:Huawei HCIE-Datacom Guide – BGP EVPN and VXLAN Routing

Comprehensive and Detailed In-Depth Explanation:

InOSPFv3, theAS-external-LSA (Type 5 LSA)is the only LSA type that can beflooded throughout the entire Autonomous System (AS). It is used to advertise external routes (routes redistributed from other routing protocols or static routes) into the OSPF domain.

B. Intra-Area-Prefix-LSA (Type 9):Used to advertise IPv6 prefixes within an OSPF area, not across the AS.

C. Link-LSA (Type 8):Advertises link-local addresses and options, only within the local link, not the entire AS.

D. Inter-Area-Prefix-LSA (Type 3 or Type 4):Advertises prefixes between OSPF areas but not throughout the entire AS.

OnlyAS-external-LSAs (Type 5)can be propagated across the whole OSPF AS, making them essential for exchanging routes learned from other routing protocols.

FEC (Forwarding Equivalence Class) in MPLS is a group of packets that are forwarded in the same manner. MPLS can classify packets into FECs based on factors such as:

Destination IP address

Source address

CoS (Class of Service)

Application type

However, Fragment Offset, which is part of the IP header used for IP fragmentation, is not used in FEC classification. Fragment offset is more relevant to how IP packets are reconstructed and not suitable or used as a classification standard for forwarding decisions in MPLS.

Exact Extract - Huawei HCIE-Datacom V1.0 MPLS Chapter:

“FEC classification may be based on various parameters including destination address, source address, and even CoS values. However, fragment offset is not used as an FEC criterion, as MPLS forwarding occurs before IP reassembly.”

Comprehensive and Detailed Explanation:

✔All four egress modesare supported inHuawei CloudCampus Virtualized Fabricto provideflexibility in connecting external networks.

✔Modes Explained:

✅Layer 3 Shared Egress→ Multiple VNs share asingle Layer 3 gatewayfor external communication.

✅Layer 2 Shared Egress→ Multiple VNs share the sameLayer 2 linkto the external network.

✅Layer 3 Exclusive Egress→ Each VN has its ownseparate Layer 3 gatewayto external networks.

✅Layer 2 Exclusive Egress→ Each VN has its owndedicated Layer 2 linkfor external connectivity.

📖Reference:Huawei HCIE Datacom – CloudCampus Fabric Egress Modes

Comprehensive and Detailed In-Depth Explanation:

InOSPF, a Not-So-Stubby Area (NSSA) is a type of area that allows for the injection ofexternal routes(like a standard area) while maintaining some stub area characteristics.

Type 7 LSAs (NSSA External LSAs)are generated by ASBRs within an NSSA.

AnABRat the NSSA boundary translates theseType 7 LSAs into Type 5 LSAsto propagate them into the rest of the OSPF domain.

This translation is necessary becauseType 5 LSAsare the standard method for advertising external routes in OSPF.

Therefore, the statement isTRUE.

Comprehensive and Detailed In-Depth Explanation:

The regular expression100.$can be broken down as follows:

100:Matches the exact sequence of digits "100".

(dot): Representsany single character.

$:Indicates theend of the line.

Therefore, the pattern100.$matches anyfour-character sequencethat starts with "100" and ends withany single character.

1000 (A): Matches because it has four digits, starting with "100" and ending with "0".

1001 (B): Matches because it has four digits, starting with "100" and ending with "1".

10000 (C): Doesnot matchbecause it hasfive digits.

100 (D): Doesnot matchbecause it only hasthree digits.

Understanding EVPN (Ethernet VPN) Label Allocation and ARP Handling

🔹What Happens in This Scenario?

1️⃣CE1 (192.168.1.1) sends an ARP request to discover the MAC address of CE2 (192.168.1.2).

2️⃣PE1 receives the ARP request and needs to forward it over the EVPN fabric.

3️⃣To send the broadcast ARP request across the EVPN network, PE1 forwards the packet to PE4 using BUM (Broadcast, Unknown unicast, and Multicast) forwarding.

4️⃣PE4, as the Designated Forwarder (DF), forwards the ARP request to CE2.

Role of Labels in EVPN VXLAN/MPLS Networks

🔹EVPN uses BGP to distribute MAC/IP reachability information using different route types.

✅EVPN Type 2 Route (MAC/IP Advertisement Route):

Used forMAC-to-IP address mappingannouncements.

Ensuresremote VTEPs (PE routers) learn MAC-to-IP bindingsdynamically.

✅EVPN Type 3 Route (Inclusive Multicast Route):

Used forBUM traffic (Broadcast, Unknown Unicast, Multicast).

Assigns a multicast label (e.g., 201) to be used for forwarding broadcast packets like ARP requests.

🔹In this case:

Thelabel 201is used forBUM traffic forwarding.

BUM traffic is handled using an EVPN Type 3 route.

Why is the Answer 3?

✅EVPN Type 3 routes carry labels for BUM traffic, which includes ARP requests.

✅Label 201 is used for multicast distribution, which is managed by Type 3 routes.

Real-World Application:

VXLAN Data Centers:UsesEVPN Type 3 routes to control BUM traffic efficiently.

Campus Networks:Preventsunnecessary flooding of ARP requests across VXLAN domains.

✅Reference:Huawei HCIE-Datacom Guide – EVPN BGP Route Types and Multicast Labeling

In Huawei’sCloudCampus Solution, VNs (Virtual Networks) areisolated by defaultfor security purposes. To enableinter-VN communication, you must explicitly configure policies that allow traffic between different VNs.

The correct way to control and enable this communication is viarouting policies, especiallyinter-VN routing policieson thecore or border nodes.

Static routesorOSPF deploymentsare not sufficient or appropriate for policy-controlled VN communication in this context.

Correct answer: D. Deploy a routing policy

Comprehensive and Detailed Explanation:

✔Hub-Spoke networkingis the mostefficient and scalablemodel for interconnecting100+ branchesin asmall and midsize campus network.

✔Why Hub-Spoke?

Centralized Control:Ahub routermanages communication betweenspoke routers(branch sites).

Reduced Complexity:Unlikefull-mesh, which requiresN(N-1)/2* connections,hub-spoke minimizes link costs.

Easier Management:Policy enforcement and security are centralized at thehub site.

✔Why Not Other Options?

❌A (Direct Networking)→ Suitable forsmall networks, but inefficient for100+ branches.

❌B (Full-Mesh Networking)→ Highcost and complexity, requiringtoo many links.

❌D (Partial-Mesh Networking)→ More scalable thanfull-mesh, buthub-spoke is more optimizedfor large branch networks.

📖Reference:Huawei HCIE Datacom – WAN Interconnection Models

DiffServ QoS model operates on per-hop behavior (PHB) and has these limitations:

A. ✅ Single-hop behavior; no end-to-end path awareness.

B. ✅ Cannot guarantee dedicated resources to individual users.

C. ✅ Does not address traffic bursts; congestion is still possible.

D. ✅ Queue limitations hinder fine-grained SLA assurance and deterministic delay.

Exact Extract – HCIE-Datacom QoS Section:

“DiffServ provides PHB without path-level control. It cannot guarantee SLA for individual users due to limited queuing mechanisms and lack of end-to-end visibility.”