H12-921-V1-0 Free Sample Questions

A financial services company is deploying a new data center using Huawei's CloudFabric solution with a BGP EVPN VXLAN fabric. The design requires that tenant VRFs be extended across multiple PoDs (Points of Delivery) for active-active services. To achieve this, a network architect plans to use EVPN Type-5 routes to advertise external prefixes into the fabric. Which component is responsible for originating these Type-5 routes and leaking them between tenant VRFs and the underlay routing table?

During the deployment of an intelligent lossless network for a new AI training cluster, a network engineer observes that while Priority Flow Control (PFC) is enabled for the RoCEv2 traffic class, congestion is still causing high latency. Further investigation reveals that Explicit Congestion Notification (ECN) is not being correctly processed. Which two components must be correctly configured to work in tandem for a Data Center Quantized Congestion Notification (DCQCN) algorithm to function effectively? (Select TWO)

A cloud service provider is using Huawei CloudFabric to offer multi-tenant services on an OpenStack platform. A new tenant requires a three-tier application (Web, App, DB) with strict security policies enforced by a virtual firewall. The network administrator needs to automate the deployment of this service chain. Which iMaster NCE-Fabric feature should be used to define the traffic path and ensure that all inter-tier traffic is steered through the virtual firewall service function?

An architect is designing the underlay network for a new spine-leaf data center fabric that will host over 2000 servers. The primary requirements are fast convergence, scalability, and operational simplicity. The fabric will use BGP EVPN for the overlay. Which underlay routing protocol would be the best practice for this scenario?

A network operations team is using iMaster NCE-Fabric for automated O&M. They receive an alert from FabricInsight indicating high network latency for a specific application. To perform root cause analysis, the team needs to visualize the end-to-end path of the application's traffic flow, including all physical and virtual nodes. Which FabricInsight feature provides this capability?

True or False: In a Huawei CloudFabric solution, microsegmentation is achieved by defining Endpoint Groups (EPGs) and then applying contracts between them. A contract is required even for EPGs within the same tenant VRF to allow traffic.

A large e-commerce platform runs its application on a Kubernetes cluster integrated with a Huawei CloudFabric data center network. To expose their services to the internet, they use the Kubernetes `LoadBalancer` service type. Which two of the following components work together to automatically provision an external IP and configure the necessary network policies on the fabric? (Select TWO)

**Case Study:** A retail company is migrating its on-premises infrastructure to a new data center built with the Huawei CloudFabric solution. The company has two primary data centers, DC1 (Primary) and DC2 (Disaster Recovery), located in different cities. The business requires active-active access to applications hosted in both data centers and seamless VM mobility for maintenance and disaster avoidance. The infrastructure consists of a VXLAN EVPN fabric in each DC, managed by iMaster NCE-Fabric. The network team is responsible for designing the Data Center Interconnect (DCI) solution. A key requirement is that a VM's IP address and MAC address must remain the same when it is moved from DC1 to DC2. Furthermore, traffic from a VM in DC1 destined for a VM in DC2 should take the most optimal path without hairpinning through a centralized gateway. To meet these requirements, the network team needs to extend the L2 and L3 domains across both data centers. They are considering different VXLAN gateway deployment models for the DCI. Which DCI design using BGP EVPN best fulfills all the stated requirements for seamless mobility and optimal traffic forwarding?

A network architect is in the capacity planning phase for a new data center. The design uses a spine-leaf topology. The leaf switches have 48x25GE server-facing ports and 8x100GE uplink ports. To avoid performance bottlenecks, the architect needs to determine the oversubscription ratio from the leaf switches to the spine switches. What is the correct oversubscription ratio for this leaf switch configuration?

A DevOps engineer is attempting to automate the creation of a new tenant VRF and associated subnets in a Huawei CloudFabric environment using RESTful APIs provided by iMaster NCE-Fabric. The engineer sends a POST request to the correct API endpoint but receives a `401 Unauthorized` error. What is the most likely cause of this issue?