The HCIE-Datacom V1.0 (H12-891_V1.0)

Segment Routing over IPv6 (SRv6) defines several headend behaviors for how SRH (Segment Routing Headers) are added:

B. H.Insert — Inserts an SRH into an existing IPv6 packet.

C. H.Encaps — Encapsulates the original packet in a new IPv6 packet with an SRH.

D. H.Insert.Red — Performs insert behavior and reduces the payload for optimization.

A. H.Encaps.Red is not a defined SRv6 behavior — it’s an invalid option.

Exact Extract - Huawei SRv6 White Paper:

“H.Insert, H.Insert.Red, and H.Encaps are valid behaviors for injecting SRH. H.Encaps.Red is not defined in the SRv6 headend behavior set.”

The incorrect statement is Option C, which says “A site can only belong to only one VPN.” This is false because a single site can belong to multiple VPNs if needed. For example, in some enterprise designs, a site may be part of both the Finance VPN and the Management VPN by configuring multiple routing instances (VRFs) and policies accordingly.

Exact Extract - HCIE-Datacom VPN Technologies Chapter:

“A site may be associated with one or more VPNs, depending on business requirements. This is achieved through VRF and route-target configurations.”

Comprehensive and Detailed Explanation:

✔ gRPC supports four types of RPCs:

Unary RPC → Client sends one request, server responds once.

Server streaming RPC → Server sends multiple responses.

Client streaming RPC → Client sends multiple requests.

Bidirectional streaming RPC → Both sides exchange multiple messages.

✔ Why B and C Are Incorrect:

❌ (B) Incorrect → The method rpc dataPublish(stream serviceArgs) returns(stream serviceArgs) {} violates gRPC syntax.

❌ (C) Incorrect → rpc stream LotsOfGreetings(HelloRequest) returns (HelloResponse) {} contains incorrect stream positioning.

???? Reference: Huawei HCIE Datacom – gRPC API and Streaming Mechanisms

Understanding VXLAN and ARP Suppression in BGP EVPN

VXLAN (Virtual Extensible LAN) is used to extend Layer 2 networks over Layer 3 infrastructure.

VXLAN uses VTEPs (VXLAN Tunnel Endpoints) to encapsulate and forward Layer 2 traffic.

BGP EVPN (Ethernet VPN) is used for MAC/IP address learning and distribution across VXLAN networks.

How ARP Request Forwarding Works in This Scenario?

1️⃣ ARP Broadcast Suppression Enabled (VXLAN Optimized Forwarding)

Since ARP broadcast suppression is enabled on VTEP1, the ARP request is not flooded.

Instead, VTEP1 checks its EVPN learned database to find PC2’s MAC address (MAC C).

2️⃣ VTEP1 Encapsulates the ARP Request

VTEP1 encapsulates the ARP request in a VXLAN packet and sends it to VTEP2.

The original ARP frame inside the VXLAN tunnel still has:

Source MAC: MAC A (PC1 ' s MAC)

Destination MAC: MAC C (PC2 ' s MAC, learned via EVPN)

3️⃣ VXLAN Outer Headers

The outer VXLAN header uses:

Source IP: 1.1.1.1 (VTEP1)

Destination IP: 2.2.2.2 (VTEP2)

VNI (VXLAN Network Identifier): 200 (BD 20)

4️⃣ VTEP2 Receives and Forwards the ARP Request

VTEP2 decapsulates the VXLAN packet and forwards the original ARP request to PC2 (MAC C).

This ensures that only the intended receiver (PC2) gets the ARP request, preventing unnecessary broadcast flooding.

Why the Answer is MAC C?

✅ B. MAC C is Correct

The destination MAC of the inner data frame is PC2’s MAC (MAC C).

This is because VTEP1 forwards the ARP request directly to VTEP2, which then forwards it to PC2.

Why Other Answers Are Incorrect?

❌ A. MAC B (Incorrect)

MAC B is VTEP1’s MAC address, which is used in the outer VXLAN header, not in the inner Ethernet frame.

❌ C. MAC A (Incorrect)

MAC A is the source MAC of the ARP request (PC1), not the destination.

❌ D. MAC D (Incorrect)

MAC D is not mentioned in the topology, so it is irrelevant.

???? Reference: Huawei HCIE Datacom – VXLAN BGP EVPN and ARP Suppression

To configure an SR-MPLS TE tunnel based on OSPF, all four options are mandatory:

A. Configure the LSR ID – LSR ID is required as the unique identifier of the router in MPLS and TE.

B. Enable segment routing in the OSPF process view – Needed to advertise segment information (SID).

C. Enable the opaque capability – Opaque LSAs are used to carry SR and TE information in OSPF.

D. Enable MPLS TE in the OSPF process view – Ensures that TE-related attributes (e.g., bandwidth) are advertised in OSPF LSAs.

Exact Extract - Huawei HCIE-Datacom SR-MPLS TE Configuration Guide:

“To support SR-MPLS TE over OSPF, it is necessary to configure the LSR ID, enable SR and MPLS TE capabilities in OSPF, and activate the OSPF opaque capability for carrying TE and SR attributes.”

Comprehensive and Detailed In-Depth Explanation:

The Strict Priority (SP) scheduling algorithm always gives preference to high-priority queues.

A: Correct. SP scheduling may lead to queue starvation because low-priority queues might never get serviced if high-priority traffic is constant.

B: Correct. SP always schedules the highest-priority queue first.

C: Correct. Among multiple low-priority queues, if SP is combined with FIFO (First In, First Out), packets that arrive first are forwarded first.

D: Incorrect. The SP algorithm does not guarantee minimum bandwidth for lower-priority queues. It prioritizes high-priority queues regardless of bandwidth configuration, which can lead to starvation of lower-priority traffic. To ensure minimum bandwidth, a Weighted Fair Queuing (WFQ) or CBWFQ (Class-Based Weighted Fair Queuing) algorithm would be more suitable.

In BGP, community attributes help control the advertisement scope of BGP routes. Here’s what each option means:

No_Export: This well-known community attribute prevents BGP routes from being advertised to external BGP (EBGP) peers. Routes with this attribute can still be advertised to internal BGP (IBGP) peers within the same AS. This ensures that the route remains within the originating AS.

No_Advertise: This attribute means that the route should not be advertised to any BGP peer, whether internal or external.

Internet: Indicates that the route can be advertised to all BGP peers, both internal and external, with no restrictions.

No_Export_Subconfed: Similar to No_Export, but it also restricts advertisement to peers within a BGP confederation. It ' s more specific and not generally used just to keep routes inside one AS.

Therefore, the correct answer is B. No_Export, as it is the attribute that ensures BGP routes remain within the AS by not being sent to EBGP peers.

Understanding Performance Measurement in Huawei CloudWAN

Huawei CloudWAN enables intelligent traffic management, real-time network monitoring, and automated operations.

???? Traditional Performance Monitoring Methods

TWAMP (Two-Way Active Measurement Protocol) → Measures latency and jitter by injecting test traffic.

Telemetry → Provides real-time network visibility, but does not measure actual service traffic performance.

BGP FlowSpec → Used for DDoS mitigation, not performance measurement.

???? iFIT (In-situ Flow Information Telemetry)

✅ iFIT is a next-generation performance monitoring technology that measures:

Real service traffic performance (latency, jitter, packet loss) in real time.

Packet forwarding behavior at each hop without injecting synthetic traffic.

Path delay and service degradation detection.

Why is Answer D Correct?

✅ iFIT provides real-time network performance monitoring based on actual service traffic.

✅ Unlike TWAMP, iFIT does not require probe packets, reducing overhead.

Real-World Application:

5G Transport Networks: Ensures low-latency paths for ultra-reliable services.

CloudWAN Optimization: Detects network congestion without affecting user traffic.

✅ Reference: Huawei HCIE-Datacom Guide – iFIT and Intelligent Performance Monitoring

In MPLS LDP, there are two types of sessions:

Direct LDP sessions: Established between directly connected LSRs using LDP Hellos over a link.

Remote LDP sessions (also called targeted LDP sessions): Established between non-directly connected LSRs. These require manual configuration of the remote peer.

In this scenario, SWA and SWC are not directly connected, so a remote (targeted) LDP session must be set up.

For a remote LDP session:

You must configure the remote peer using the command:

[SWA-MPLS-LDP] targeted-peer x.x.x.x

Where x.x.x.x is the LSR ID of SWC (33.3.3.3 in this case).

The current configuration does not include this required step.

Thus, the correct statement is:

D. You need to create a remote peer and specify an LSR ID for the remote peer.