In an EVPN multihoming scenario, a CE device is connected to multiple PEs within the same Ethernet Segment. When the CE sends Broadcast, Unknown Unicast, or Multicast (BUM) traffic, all connected PEs receive it. Without a control mechanism, each PE would flood the packet into the EVPN core. This would cause remote PEs to receive multiple copies of the same packet, which they would then forward to their connected CEs. The Designated Forwarder (DF) election process ensures that only one PE (the DF) is responsible for forwarding BUM traffic from the CE, thus preventing the delivery of duplicate packets to remote hosts.

A. Loop prevention between PEs in the EVPN core is primarily handled by split-horizon rules, not the DF election for CE-originated traffic.

B. MAC addresses are learned and distributed via the BGP control plane (EVPN Type-2 routes), a process separate from the DF's data plane forwarding role for BUM traffic.

C. While duplicate packets increase traffic, the specific problem solved by the DF is duplication itself, not necessarily a self-propagating traffic storm. "Duplicate packets" is the most precise answer.

1. IETF. (2015). RFC 7432: BGP MPLS-Based Ethernet VPN. Section 8.5, "Forwarding BUM Traffic," states, "In order to prevent duplicate packets from being delivered to a CE, only one of the PEs attached to a given Ethernet segment is allowed to forward BUM traffic... This PE is called the Designated Forwarder (DF)." (p. 20). Retrieved from https://doi.org/10.17487/RFC7432

2. Juniper Networks TechLibrary. (2023). EVPN Multihoming Overview. "For a given Ethernet segment, only one of the PE routers, called the designated forwarder (DF), forwards BUM traffic to the CE. This restriction prevents duplication of BUM packets." Retrieved from https://www.juniper.net/documentation/us/en/software/junos/evpn-vxlan/topics/concept/evpn-multihoming-overview.html

Juniper's Zero Touch Provisioning (ZTP) process can fetch and execute a script to perform advanced device configuration. The Junos OS environment on ZTP-capable devices natively supports the execution of scripts written in standard UNIX shell scripting language, Python, and SLAX (Style Sheet Language Alternative Syntax). These scripts are executed directly on the device after it boots and receives the script location from a DHCP server, enabling powerful and flexible automated provisioning beyond a static configuration file.

A. The Junos ZTP process supports Shell, JScript, and Ansible: Ansible is an external automation tool, not a script language executed on the device during ZTP. JScript is supported for commit scripts, not typically for ZTP.

B. The Junos ZTP process supports Python, SLAX, and Perl: Perl is not a natively supported scripting language for on-box execution in the Junos ZTP context.

C. The Junos ZTP process supports JScript, Ansible, and Perl: This is incorrect for the same reasons as options A and B; Ansible, Perl, and JScript are not the correct languages for this process.

1. Juniper Networks TechLibrary. (2023). Automating Junos Device Bootstrapping Using ZTP. "The script can be a shell script (for UNIX) or a SLAX or Python script. You can use the script to generate a device-specific configuration file, and then load the configuration file onto the device."

2. Juniper Networks TechLibrary. (2023). Junos OS Automation Scripting User Guide. Introduction. This guide details the supported on-box scripting languages, highlighting SLAX and Python as primary choices for automation scripts that interact with the Junos OS.

The output of the show evpn database command provides details about MAC and IP addresses learned within the EVPN instance. For the MAC address 40:00:dc:01:00:04, the IP address column clearly shows 10.4.4.5. The VLAN-ID column shows 110002, which corresponds to the Virtual Network Identifier (VNI), indicating the host is part of VN110002. This information is learned locally and advertised to other PEs via EVPN Type-2 (MAC/IP Advertisement) routes.

B. The Interface column shows ae1.0, the physical interface where the host is connected. The IRB interface is a logical L3 interface, not the source of the MAC learning.

D. The Source column shows 00:00:00:00:00:00, which indicates a locally learned, single-homed host. A host originating from an ESI LAG would show a non-zero ESI value here.

1. Juniper Networks Documentation

"show evpn database". This command display output fields are detailed

confirming that the IP address and VLAN-ID columns map directly to the host's IP and VNI. The Source field is documented to show the ESI for multi-homed devices.

2. Juniper Networks TechLibrary

"EVPN-VXLAN Data Center Interconnect". In the section on DCI Operations

examples of show evpn database illustrate how local MAC/IP bindings are displayed with their corresponding VNI (VLAN-ID) and interface.

When comparing VXLAN control planes, EVPN (Ethernet VPN) offers significant advantages over a multicast-based approach. EVPN uses BGP to distribute MAC address and VTEP reachability information, which is highly scalable, enables fast convergence upon network changes, and supports advanced features like MAC mobility. In contrast, a multicast control plane relies on data-plane flood-and-learn, which requires manual mapping of VNIs to multicast groups and is less agile and scalable. While both methods ultimately facilitate MAC learning (EVPN via control-plane and multicast via data-plane flooding), EVPN's approach is far more efficient for modern data centers.

A. EVPN is inefficient and does not scale well.

This is false. EVPN was specifically designed to overcome the scaling and efficiency limitations of older technologies like VPLS and multicast-based VXLAN.

E. Multicast does not require as many resources.

This is misleading. While it avoids a BGP control plane, it requires PIM state and multicast forwarding entries on all VTEPs, and the associated flooding can consume significant bandwidth.

1. Juniper Networks Documentation, EVPN User Guide, "EVPN-VXLAN Overview". "EVPN provides a scalable, efficient, and robust solution for Layer 2 and Layer 3 connectivity in a data center... Compared to using a multicast model for BUM traffic forwarding, EVPN is more efficient."

2. Juniper Networks Documentation, Understanding EVPN with VXLAN Data Plane Encapsulation. "In a multicast-based VXLAN deployment, you must manually configure the VNI-to-multicast-group mapping on each VTEP. This approach is not agile... EVPN solves these issues by providing a control plane that advertises MAC addresses."

3. Juniper Networks White Paper, Data Center Fabric with EVPN and VXLAN. "EVPN provides a control plane for VXLAN that allows VTEPs to learn about remote MAC addresses and VTEPs before they need to forward traffic... This control plane approach is much more scalable than the data plane flood-and-learn approach used with multicast."

Unnumbered BGP relies on IPv6 link-local addresses for establishing BGP sessions, requiring family inet6 configuration on interfaces. When configured, it automatically provisions IPv6 peering using these link-local addresses. This approach is used in modern data center fabrics to simplify configuration while maintaining scalability. The BGP session itself runs over IPv6, though it can advertise both IPv4 and IPv6 routes using appropriate address families.

C. family inet (IPv4) is not required for unnumbered BGP operation.

D. Unnumbered BGP establishes IPv6 peering, not IPv4 peering automatically.

1. Juniper Networks Day One: Data Center Fundamentals, Chapter on IP Fabric Design, pages 89-92

2. Juniper Technical Documentation: "BGP Unnumbered Interfaces" (TechLibrary Article ID: KB32543)

3. RFC 5549: "Advertising IPv4 Network Layer Reachability Information with an IPv6 Next Hop"

4. Juniper Networks JNCIS-DC Study Guide, Section 4.3: "IP Fabric with EBGP"

5. Juniper Configuration Guide: "Understanding BGP Unnumbered Interfaces in IP Fabrics" (2023.2 Release)

The exhibit illustrates an Edge-Routed Bridging (ERB) architecture, also known as a leaf-routed or ToR-routed model. The defining characteristic of this design is the placement of the Layer 3 gateway functionality on the leaf switches. As shown in the diagram, "L3 Gateway" is present on Leaf 1 and Leaf 2. This allows for inter-VLAN routing to occur at the edge of the fabric, directly on the ingress leaf device. This optimizes east-west traffic patterns within the data center by preventing traffic from having to traverse to the spine layer for routing.

A. The diagram lacks the specific details of MAC-in-MAC encapsulation that would be required to identify it as a PBB-EVPN architecture.

C. Over-The-Top (OTT) is a general term for an overlay network; it does not describe the specific routing architecture (i.e., where the L3 gateway resides) like ERB or CRB do.

D. In a Centrally-Routed Bridging (CRB) architecture, the L3 gateway is located on the spine devices, and the leaf devices perform only Layer 2 functions. The exhibit clearly shows L3 gateways on the leafs.

1. Juniper Networks. (2021). Day One: Data Center Fabric with EVPN and VXLAN. Chapter 2

"EVPN-VXLAN Architecture"

Section "Edge-Routed Bridging Overlay". This section states

"In an ERB overlay

the VXLAN L3 gateway is implemented on the leaf devices... This model is also known as a host-based overlay or a top-of-rack (ToR) based overlay." This directly corresponds to the provided exhibit.

2. Juniper Networks TechLibrary. (2023). EVPN-VXLAN Design and Implementation Details Guide. Section: "EVPN-VXLAN Fabric Architecture Models". This guide contrasts the two primary models: "In an edge-routed bridging (ERB) overlay

the Layer 3 gateway is on the leaf devices... In a centrally-routed bridging (CRB) overlay

the Layer 3 gateway is on the spine devices." The exhibit aligns with the ERB definition.

3. Juniper Networks TechLibrary. (2024). Understanding EVPN-VXLAN Edge-Routed Bridging Overlay Model. This document explicitly defines the architecture: "In an edge-routed bridging overlay

the default Layer 3 gateway resides on the leaf devices. The leaf devices

therefore

provide both Layer 2 and Layer 3 forwarding." This confirms the analysis of the diagram.

The scenario requires a high-performance network for AI workloads without any Layer 2 VLAN requirements. An IP fabric using External BGP (EBGP) is a pure Layer 3 architecture that perfectly fits this need. In this design, each link between devices (e.g., leaf-to-spine) is a routed point-to-point link, and BGP is used to exchange reachability information for server endpoints. This approach is highly scalable, stable, and avoids the complexities of Layer 2 protocols and overlay technologies like VXLAN, which are unnecessary when L2 adjacency is not required.

A. an IP fabric using PIM-SM to signal VXLAN overlay: VXLAN is an overlay technology designed to extend Layer 2 segments, which is explicitly not a requirement.

B. an IP fabric using the EVPN-MPLS architecture: EVPN-MPLS is primarily used for extending Layer 2 and Layer 3 VPN services, which adds unnecessary complexity when only L3 connectivity is needed.

C. an IP fabric with an EVPN-VXLAN architecture: This is a common data center architecture, but its main purpose is to provide Layer 2 overlay services, which contradicts the stated requirement.

1. Juniper Networks TechLibrary. (2023). IP Fabric Underlay Network Design and Implementation. "An IP fabric is a simple, scalable, and resilient architecture that is built on a Layer 3-only infrastructure... In an IP fabric, we recommend using EBGP as the routing protocol between the leaf and spine devices."

2. Juniper Networks Day One+. (2022). Data Center Fundamentals. Chapter 3: "Data Center Architectures", Section: "IP Fabric". This section describes the IP Fabric as a two-tier spine-and-leaf architecture using BGP for routing, ideal for modern data centers that do not require L2 extension.

In a typical BGP-based IP fabric (Clos architecture), leaf and spine switches are directly connected. A single point-to-point EBGP session is established over each of these links, which is a simple and scalable design. If each device in the fabric uses a unique AS number (a common design choice), a leaf switch will learn routes to a given destination from multiple spines, each with a different AS path. To enable Equal-Cost Multi-Path (ECMP) and load-balance traffic across all spines, the multipath multiple-as statement is required on the leaf switches. This command allows BGP to consider paths through different autonomous systems as equal for multipathing.

A. If all leaf devices use the same AS, BGP loop prevention will cause routes to be dropped unless as-override or allow-as in is configured.

D. EBGP is the primary routing protocol used for reachability within the IP fabric (between leaf and spine nodes), not just for external connectivity.

1. Juniper Networks TechLibrary. (2024). EVPN-VXLAN Campus Architectures. "Configure BGP in the Core and Distribution Layers". The guide states, "To enable load balancing, include the multipath multiple-as statement... This statement allows the device to load-balance traffic across multiple BGP paths that are in different autonomous systems (ASs)."

2. Juniper Networks TechLibrary. (2024). BGP User Guide. "Examples: Configuring BGP for an IP Fabric". The examples consistently show a single EBGP session between directly connected leaf and spine peers and the use of multipath multiple-as for ECMP.

The mac-vrf routing-instance type is used for large-scale, multi-tenant EVPN deployments, often where overlapping VLANs are required. This instance type mandates the use of a "service-provider style" interface configuration on host-facing ports. This involves configuring logical units under the physical interface with specific VLAN tagging and encapsulation (e.g., encapsulation vlan-bridge). Additionally, mac-vrf instances support different service models, including vlan-based, where each VLAN is mapped to its own unique bridge domain and VNI, providing traffic separation.

B. Host-facing interfaces must be configured using enterprise-style configuration.

Enterprise-style configuration (interface-mode trunk, vlan members) is used with virtual-switch instances, not mac-vrf.

C. Spine-facing interfaces must be configured using an enterprise-style configuration.

Spine-facing interfaces are part of the IP underlay and are configured as standard L3 interfaces, not for VLAN services.

1. Juniper Networks TechLibrary, "Configuring EVPN with a MAC-VRF Routing Instance," EVPN Feature Guide. This guide explicitly states, "For each MAC-VRF instance, you must configure the instance-type mac-vrf statement...On the interfaces that you configure in the MAC-VRF instance, you must use a service-provider-style configuration."

2. Juniper Networks TechLibrary, "service-type (MAC-VRF)," EVPN Feature Guide. This page documents the available options for the service-type statement within a mac-vrf instance, which include vlan-based and vlan-aware.

The requirement is to selectively redirect a specific traffic type within a tenant's network to a service (like a firewall) without affecting all other traffic. Filter-based forwarding (FBF) is the ideal solution. FBF uses a firewall filter to match specific traffic based on criteria like protocol or port numbers. When traffic matches the filter, an action redirects it to a different next-hop or routing instance that points to the firewall. All other non-matching traffic is forwarded normally within the tenant's VRF, satisfying all requirements.

A. Route leaking is used to share routes between different VRFs, not to steer traffic based on its type within a single VRF.

C. Placing the server on a unique subnet only provides L3 separation; it does not inherently filter or redirect specific traffic types.

D. A static route would redirect all traffic destined for the server to the firewall, not just a specific traffic type as required.

1. Juniper Networks TechLibrary. (2023). Filter-Based Forwarding Overview. "Filter-based forwarding (FBF) provides a mechanism to forward packets based on a configured filter... You can use FBF to route traffic to specific destinations or to policy-based routing." Retrieved from https://www.juniper.net/documentation/us/en/software/junos/routing-policy/topics/concept/filter-based-forwarding-overview.html

2. Juniper Networks TechLibrary. (2023). Example: Configuring Filter-Based Forwarding for Specific Traffic. This document provides a direct example of using a firewall filter to match traffic and forward it to a different routing instance, which is the core of the solution. Retrieved from https://www.juniper.net/documentation/us/en/software/junos/routing-policy/topics/example/filter-based-forwarding-example.html