JN0-281 Data Center Associate (JNCIA-DC) Questions and Answers

To advertise OSPF-learned routes to BGP neighbors on Junos, you must explicitly control route redistribution using policy. Junos does not automatically redistribute routes between routing protocols simply because both protocols are enabled. Instead, you create a routing policy that matches the routes you intend to export, in this case routes whose protocol origin is OSPF. The policy must accept those matched routes so they become eligible for advertisement.

After the policy is created, you must apply it as an export policy under the relevant BGP group or under the BGP protocol hierarchy, depending on your design. Export policy controls what your router sends to BGP peers. When the export policy accepts OSPF routes, Junos advertises those routes to the BGP neighbors in that group, subject to any additional BGP constraints such as next-hop handling, route families, and any peer-specific policy terms.

Creating a policy to match and accept BGP routes is not required for this goal because that would influence what routes are imported into your routing table from BGP, not what you export. Likewise, applying a policy as a BGP import policy affects received routes, not OSPF-to-BGP redistribution. In data center fabrics, this policy-driven approach prevents unintended route leakage and keeps redistribution tightly scoped to the prefixes that must be carried between domains.

The exhibit shows a Folded IP Clos Architecture, which is also referred to as a 3-stage Clos network design. This architecture typically consists of two layers of switches:

Spine Layer: The top row of switches.

Leaf Layer: The bottom row of switches.

Step-by-Step Breakdown:

Clos Architecture:A 3-stage Clos network has two types of devices: spine and leaf. In this design, each leaf switch connects to every spine switch, providing a high level of redundancy and load balancing.

Stage Explanation:

Stage 1: The first set of leaf switches.

Stage 2: The spine switches.

Stage 3: The second set of leaf switches.

The Folded Clos architecture shown here effectively "folds" the 3-stage design by combining the ingress and egress leaf layers into one, reducing it to two visible layers, but still maintaining the overall 3-stage architecture.

Juniper Reference:

IP Clos Architecture: The 3-stage Clos design is commonly used in modern data centers for high availability, redundancy, and scalability.

To confirm that your router is sending a specific route to a particular BGP neighbor, you must inspect the outbound advertisements toward that neighbor. In Junos, the command that shows routes being advertised to a peer is show route advertising-protocol bgp with the neighbor address specified. Adding 0.0.0.0 to the command filters the output to the default route, making it the most direct way to validate that the default route is actually being exported to peer 10.100.25.6. This is especially important in data center deployments where default route advertisement is often controlled by policy, conditional origination, or specific export terms, and where you want to verify the real operational result rather than just configuration intent.

The receive-protocol variant shows what you are learning from that neighbor, not what you are sending to it. The show route protocol bgp 0.0.0.0 command only confirms that the default route exists in your local routing table as a BGP-learned route, which does not prove it is being exported to the neighbor. The show route protocol static 0.0.0.0 command would confirm the presence of a static default route locally, but again it does not confirm that it is being advertised over BGP. Therefore, the outbound advertisement command is the correct verification method.

In Junos OSPF, the passive setting is used when you want the prefix on an interface to be advertised into OSPF, but you do not want the interface to form OSPF neighbor adjacencies or send OSPF Hello packets. This is a common data center requirement for server-facing, management, or loopback-related interfaces where you want the subnet to be reachable through the routing domain, but there is no OSPF-capable neighbor on that link. By making the interface passive, the router still includes the connected network in OSPF as an internal route, so other routers learn how to reach that subnet. At the same time, the interface does not participate in adjacency formation, which reduces unnecessary protocol chatter and eliminates the risk of accidentally forming neighbors on an untrusted or unintended segment.

The other options do not meet the requirement. Interface-type changes the OSPF network type behavior, such as broadcast or point-to-point, but it does not stop adjacency formation while still advertising the network. Strict-bfd relates to failure detection behavior when BFD is used with routing protocols and does not control whether OSPF forms neighbors. Link-protection is related to fast reroute and protection mechanisms, not suppressing OSPF adjacency on an interface.

Nonstop active routing is the Junos high availability capability that focuses on preserving routing protocol operation and routing information across a Routing Engine switchover. In platforms with redundant Routing Engines, a failure of the primary Routing Engine can otherwise reset routing protocol processes, tear down adjacencies, and trigger reconvergence. NSR mitigates this by synchronizing routing protocol state so that the backup Routing Engine can continue routing protocol operations with minimal disruption. This includes maintaining protocol session continuity and keeping the routing information base stable, which directly protects traffic that depends on those routes.

In data center environments, this is particularly important for routed fabrics where BGP or OSPF underlay reachability supports overlay services and east west application traffic. By keeping routing information consistent during the control-plane transition, NSR reduces route churn and helps avoid transient blackholing or microbursts caused by reconvergence.

GRES is closely related but addresses a different scope. GRES helps the forwarding plane continue forwarding during a Routing Engine switchover by preserving certain system and interface states. However, GRES alone does not guarantee that routing protocol sessions and routing information remain uninterrupted. BFD and LACP are valuable availability tools, but they are not Routing Engine redundancy features and do not preserve routing state during a Routing Engine failover.

EBGP is defined as BGP peering between different autonomous systems, which makes statement A correct. In data center IP fabrics, it is common to assign different private AS numbers to leaves and spines or to use a structured AS design so that every leaf forms EBGP sessions to each spine. This provides clear policy boundaries, straightforward troubleshooting, and predictable route propagation without needing an additional interior gateway protocol to carry underlay reachability.

Statement C is also correct because EBGP can be deployed without a supporting IGP. BGP itself can distribute the underlay routes needed for fabric reachability, such as loopback addresses and point to point link prefixes. This is a widely used approach for leaf spine fabrics because it reduces protocol complexity and avoids running multiple control planes for the underlay. Convergence can be improved using multipath, rapid failure detection mechanisms, and consistent routing policy.

Statement B is incorrect because BGP within a single AS is IBGP, not EBGP. Statement D is incorrect because while some designs may choose to run an IGP and use BGP only for certain functions, EBGP does not inherently require an IGP to operate or to provide underlay connectivity in a fabric design.

On Junos-based Ethernet switching platforms used in data centers, Layer 2 forwarding is handled primarily in the forwarding plane. When a switch receives a frame, it first performs source MAC learning. If the source MAC address is not already present in the Ethernet switching table for that VLAN or bridge domain, the switch creates a new entry that maps the source MAC to the ingress interface and the associated VLAN context. This learning step is fundamental to building the MAC table dynamically and enables efficient forwarding for subsequent frames destined back to that source.

Next, the switch attempts to forward the frame based on the destination MAC lookup in the same VLAN or bridge domain. Because the destination MAC is also not in the Ethernet switching table, the frame is treated as an unknown unicast. The default behavior for unknown unicast in a Layer 2 broadcast domain is to flood the frame out all other interfaces that belong to that VLAN or bridge domain, excluding the ingress interface. Flooding ensures the frame has the best chance of reaching the correct destination host. When the destination responds, the switch then learns that MAC address as a source on the return traffic, allowing future traffic to be forwarded as known unicast instead of flooded.

The switch does not drop the frame by default, and it does not forward the frame to the Routing Engine because this is normal Layer 2 bridging behavior, not a control-plane routing decision.

On Junos switching platforms used in data centers, a VLAN is a Layer 2 broadcast domain. To make a VLAN functional for user traffic, you define the VLAN with a name and typically a VLAN ID, and you associate Layer 2 interfaces with that VLAN so frames arriving on those interfaces are placed into the correct broadcast domain. Without interface membership, the VLAN exists as configuration but does not carry endpoint traffic because no ports participate in it. This is why assigning a VLAN name or ID and associating Layer 2 interfaces to the VLAN is a correct requirement.

Trunk mode interfaces are designed to carry multiple VLANs over a single physical link using 802.1Q tagging. In a data center, trunks are common on leaf-to-spine uplinks, switch-to-switch connections, and server connections where the host or hypervisor tags multiple VLANs. Therefore, assigning one or more VLANs to a trunk port is correct.

An IRB interface is not required for every VLAN. IRB is only needed when the VLAN requires Layer 3 gateway functionality, such as inter-VLAN routing or default gateway services for that subnet. Pure Layer 2 VLANs do not need IRB. Also, an access mode interface is intended to belong to a single VLAN and typically carries untagged traffic, so assigning multiple VLANs to an access mode interface is not correct in standard Ethernet switching behavior.

The BGP session in the exhibit shows the state as Connect, which indicates that the TCP session between the BGP peers has not been fully established.

Step-by-Step Breakdown:

BGP State "Connect":

The Connect state is the second stage in the BGP finite state machine (FSM). At this stage, BGP is trying to establish a TCP session with the peer, but the session has not yet been successfully established.

A successful TCP three-way handshake (SYN, SYN-ACK, ACK) is required before BGP can progress to the OpenSent state, where the peers exchange BGP Open messages.

Possible Causes:

A firewall blocking TCP port 179.

Incorrect IP addresses or network connectivity issues between the BGP peers.

Juniper Reference:

BGP Troubleshooting: In Junos, if a BGP session is stuck in the Connect state, the issue is likely due to a failure in establishing the underlying TCP connection.

Link aggregation groups, implemented on Junos as aggregated Ethernet interfaces, allow multiple physical Ethernet links to operate as one logical Layer 2 interface. This increases available bandwidth and provides link level resiliency because member links can fail without taking down the logical interface, as long as at least one member remains operational. In data center leaf spine designs, link aggregation is commonly used for server dual homing, uplinks to appliances, or inter switch connectivity where parallel links are desired with a single logical adjacency.

From a forwarding perspective, the device distributes traffic across member links using a hashing algorithm based on packet header fields so that individual flows remain in order while the aggregate uses multiple links. Control plane operation can be static or negotiated with LACP. With LACP, both sides exchange protocol information to ensure consistent bundling and to remove failed or miswired members automatically. This makes link aggregation a core high availability building block at the link layer, independent of Routing Engine redundancy features.

Graceful Routing Engine switchover, nonstop bridging, and nonstop active routing are control plane redundancy features. They are designed to minimize disruption during Routing Engine failover or preserve protocol state, but they do not combine multiple physical Ethernet interfaces into one logical link layer interface.

Verification sources from Juniper documentation

https://www.juniper.net/documentation/us/en/software/junos/interfaces-ethernet/topics/topic-map/understanding-lacp.html

https://www.juniper.net/documentation/us/en/software/junos/interfaces-ethernet/topics/topic-map/aggregated-ethernet-interfaces-overview.html

The diagram shows multiple autonomous systems, with each device labeled with a different AS number. Because the peers are in different autonomous systems, the correct BGP session type is external BGP. In a typical Juniper data center IP fabric underlay, this design is intentional: EBGP is used between leaf and spine devices to simplify policy boundaries, improve operational clarity, and avoid the additional mechanisms commonly required in large internal BGP designs.

In an EBGP-based underlay, neighbors are most commonly formed over the directly connected point-to-point links between leaf and spine. That means the BGP peering is established using the IP addresses configured on those physical routed interfaces, not loopback addresses. Using physically connected addresses aligns with the default EBGP behavior where the TTL is one hop, and it reduces configuration complexity because no multihop settings or additional reachability dependencies are required to bring up the session.

Peering using loopbacks is possible, but it typically requires EBGP multihop and a routing method to ensure reachability to the remote loopback before BGP can establish. That approach is more common for overlay control-plane sessions or for specific design requirements, rather than for the simplest and most common underlay implementation. Therefore, the two true statements are that the exhibit uses EBGP and that devices should peer using physically connected IP addresses.

A data center underlay IP fabric is a routed leaf-spine network designed to provide scalable Layer 3 connectivity between all fabric nodes. A key property of these fabrics is that there are multiple equal-cost paths between any two endpoints, typically across multiple spine devices. Equal-Cost Multi-Path load balancing is used to distribute traffic across those parallel paths. The routing table installs multiple next hops for the same destination prefix, and the forwarding plane selects an egress link per flow using a hash, which keeps packets in-order within a flow while using the fabric’s aggregate bandwidth. This makes statement B correct.

Because the underlay is routed, loop avoidance is handled by the routing protocol and the fundamental properties of IP forwarding, not by spanning tree. Routing protocols compute a loop-free forwarding topology and use mechanisms like shortest-path calculation and next-hop selection so that even when multiple paths exist, traffic is forwarded along valid loop-free routes. This makes statement C correct. Statement D is incorrect because spanning tree is a Layer 2 loop prevention mechanism and is not required or desired in a routed underlay fabric. Statement A is also incorrect because traffic distribution depends on ECMP and forwarding behavior, not on all devices being the same hardware model. Mixed platforms can interoperate as long as the design accounts for capacity, features, and consistent routing behavior.

In BGP (Border Gateway Protocol), when evaluating multiple routes to the same destination, the first attribute that is considered is the local preference value. The local preference is a BGP attribute used to influence outbound routing decisions within an Autonomous System (AS).

Step-by-Step Breakdown:

Local Preference:The local preference attribute is used to determine which path is preferred for traffic leaving the AS. The higher the local preference value, the more preferred the route.

BGP Path Selection:The BGP path selection process evaluates the following attributes in this order:

Local Preference (higher is preferred)

AS Path (shorter is preferred)

Origin (IGP > EGP > incomplete)

MED (Multi-Exit Discriminator) (lower is preferred)

Juniper Reference:

BGP Path Selection: In Junos, the local preference attribute is the first to be evaluated when determining the best path for outbound traffic.

EBGP is the BGP session type formed between different autonomous systems. In Juniper data center IP fabrics, EBGP is frequently used for the underlay because it can provide all required reachability without an additional interior gateway protocol. The fabric can advertise loopbacks and point-to-point link subnets directly in BGP, then use ECMP to install multiple equal-cost next hops. This is why EBGP sessions do not require an IGP as a fundamental dependency. Some designs still add an IGP for other reasons, but EBGP itself can carry the underlay routes needed for full fabric connectivity.

EBGP sessions are also typically established between directly connected peers. In a standard leaf-spine underlay, each leaf peers with each spine over the routed physical links between them, using the interface IP addresses on those point-to-point subnets. This matches EBGP default behavior, including a one-hop TTL expectation and straightforward operational troubleshooting.

Loopback peering is not required for EBGP. It is possible, but it usually needs additional configuration such as multihop and a routing method to ensure reachability to the remote loopback before the BGP session can form. EBGP also supports sessions with non-directly connected peers when multihop is configured, so it is incorrect to claim EBGP does not support that capability.

A Layer 2 switch builds its Ethernet switching table through a learning process based on the source MAC addresses of incoming frames. When a frame arrives on an interface within a VLAN or bridge domain, the switch examines the source MAC address and associates it with the ingress interface and VLAN context. If the MAC address is new, the switch creates a new entry; if it already exists but is seen on a different interface, the switch updates the entry to reflect the new location. This dynamic learning is fundamental to efficient unicast forwarding and is why option A is correct.

Once the switch has learned MAC-to-port mappings, it can forward subsequent frames destined to those MAC addresses as known unicast out the specific egress interface rather than flooding them. If the destination MAC is unknown, the switch typically floods the frame within the VLAN or bridge domain to discover the correct destination. When the destination replies, the switch learns that MAC as a source, completing the learning cycle.

Spanning Tree Protocol does not provide a MAC table and there is no concept of downloading an Ethernet switching table from a root bridge. STP’s role is loop prevention and topology control at Layer 2, not MAC learning distribution. Firewall filters can enforce policy but do not create the switching table. Recording destination MAC addresses would not correctly learn endpoint locations because the destination can be unknown when the first frames are sent; source learning is reliable because every received frame carries the sender’s MAC address.

Bidirectional Forwarding Detection is a lightweight, fast failure-detection mechanism used to quickly determine whether a forwarding path between two systems is operational. In Juniper data center networks, BFD is commonly paired with routing protocols such as BGP and OSPF to accelerate convergence. Instead of waiting for a routing protocol hold timer or dead interval to expire, BFD continuously exchanges small control packets between neighbors at a configured interval. If the local device stops receiving these control packets for a negotiated detection time, it declares the BFD session down.

This down event can then be used to immediately signal the routing protocol that the neighbor is no longer reachable, causing faster withdrawal of routes or faster reroute to alternate paths. This is critical in leaf-spine fabrics where rapid failover is expected to maintain application availability and to keep ECMP path sets current. BFD is designed to detect failures in the forwarding plane, including link failures, interface failures, or other failures that prevent packets from being successfully exchanged between the two endpoints. It is not intended to detect routing loops, and it does not specifically track route flaps as a function; flapping is a symptom that may occur when failures happen repeatedly.

Destination host unreachable messages are part of ICMP processing and are unrelated to BFD’s purpose. In short, BFD’s value is fast, protocol-independent failure detection for forwarding adjacency health.

In Ethernet switching, the bridge table is the data structure that maps MAC addresses to the switch interfaces where those MAC addresses were learned. Juniper commonly describes this function as the Ethernet switching table and also refers to it as the forwarding table in Layer 2 contexts. The concept is the same: the switch learns source MAC addresses from incoming frames, associates them with an ingress port and VLAN or bridge domain, and then uses that learned information to forward future frames to the correct egress port as known unicast.

Calling it a forwarding table is accurate because its primary operational purpose is deciding how to forward Layer 2 frames efficiently. When a destination MAC is present in the table, the switch performs a unicast forward to the learned port. When a destination MAC is not present, the switch treats it as unknown unicast and floods it within the VLAN or bridge domain, while still learning the source MAC for future use.

The term forwarding information table is more strongly associated with Layer 3 routing, where a FIB represents resolved next hops for IP prefixes in the forwarding plane. That is a different structure than the Layer 2 bridge or MAC table. The other options are not standard Juniper terms for this function.

Verification sources from Juniper documentation

https://www.juniper.net/documentation/us/en/software/junos/multicast-l2/topics/topic-map/ethernet-switching-components.html

https://www.juniper.net/documentation/us/en/software/junos/multicast-l2/topics/concept/ethernet-switching-table-understanding.html