350-601 Practice Test 40 — 15 Questions

Question 1

An engineer is troubleshooting a BGP EVPN issue where leaf switches are not learning remote MAC addresses. The underlay is reachable and BGP sessions are established. What is a likely cause?

Accepted Answer

Missing 'address-family l2vpn evpn' under the BGP neighbor configuration. Even when the underlay is reachable and BGP sessions are established, the BGP EVPN control plane requires the 'address-family l2vpn evpn' to be explicitly activated under the BGP neighbor configuration. Without this, the leaf switch will not exchange EVPN routes (Type-2 MAC/IP advertisement routes) with its peers, so remote MAC addresses are never learned. This is a common misconfiguration because BGP sessions may appear up but are not carrying the necessary EVPN address family.

Answer

NVE interface not enabled

Answer

Incorrect VNI configuration

Answer

MTU mismatch on the underlay

Question 2

An engineer is hardening a Nexus switch. Which THREE actions should be taken? (Choose three.)

Accepted Answer

Disable unused services like HTTP and Telnet. Option B is correct because disabling unused services such as HTTP and Telnet removes cleartext and unnecessary management attack surfaces, leaving only secure protocols enabled on the Nexus switch. Option C is correct because Control Plane Policing (CoPP) rate-limits and filters traffic destined to the switch's control plane (e.g., routing protocols, management traffic), protecting the supervisor from DoS and CPU exhaustion attacks. Option E is correct because enabling SSH only for remote management ensures encrypted, authenticated administrative access instead of insecure Telnet. Option A is not correct because SNMPv1 uses cleartext community strings and lacks strong authentication, so it should be avoided or replaced with SNMPv3. Option D is not correct because enabling LLDP on all interfaces is not a hardening action; LLDP advertises device and topology information and should be limited to trusted links or disabled where not needed.

Answer

Use SNMPv1 for monitoring

Answer

Enable LLDP on all interfaces

Question 3

Which two are benefits of using Cisco UCS Central for multi-domain management? (Choose two.)

Accepted Answer

Allows creation of global service profile templates.. Option C is correct because Cisco UCS Central enables administrators to define global service profile templates that can be shared and instantiated across multiple UCS domains, ensuring consistent policy and configuration deployment at scale. Option D is correct because UCS Central provides a single pane of glass, a unified management interface that lets administrators monitor and manage multiple UCS Manager-managed domains from one centralized point. Options A and B are incorrect because UCS Central operates on top of, not instead of, Fabric Interconnects and UCS Manager—each domain still requires its own Fabric Interconnects and UCS Manager instance. Option E is incorrect because UCS Central does not perform direct CIMC management of servers; CIMC management remains within each domain's UCS Manager, while UCS Central manages policies and service profiles at the domain level.

Answer

Eliminates the need for Fabric Interconnects.

Answer

Replaces the need for UCS Manager in each domain.

Answer

Enables direct CIMC management of all servers.

Question 4

Which THREE components are required to implement Fibre Channel over Ethernet (FCoE) in a data center?

Accepted Answer

Converged Network Adapter (CNA). Option B, the Converged Network Adapter (CNA), is correct because FCoE requires a hardware adapter that combines a standard Ethernet NIC and a Fibre Channel HBA, encapsulating FC frames into Ethernet frames while offloading FCoE and FIP processing. Option D, the FCoE Initialization Protocol (FIP), is correct because FIP is the protocol used to discover FCFs (FCoE Forwarders), perform fabric login (FLOGI), and maintain virtual links between ENodes and the FC fabric over Ethernet. Option E, Data Center Bridging (DCB)-capable switches, is correct because FCoE depends on DCB features such as Priority Flow Control (PFC, 802.1Qbb) and Enhanced Transmission Selection (ETS, 802.1Qaz) to provide lossless, class-based Ethernet transport for storage traffic. Option A is not required because FCoE uses Ethernet switches with FCF capability rather than a traditional Fibre Channel switch with an E-port, which is used for FC-to-FC ISL connections. Option C is not required because iSCSI initiator software is used for iSCSI over TCP/IP, not for FCoE, which encapsulates native Fibre Channel frames rather than SCSI over TCP.

Answer

Fibre Channel switch with E-port

Answer

iSCSI initiator software

Question 5

Which three components are part of the ACI Management Information Tree (MIT)? (Choose three.)

Accepted Answer

Tenant. In Cisco ACI, the Management Information Tree (MIT) is the hierarchical object model that stores all managed objects, and it is organized under the top-level 'uni' and 'polUni' containers. Option B (Tenant) is correct because a tenant (fvTenant) is a top-level policy container directly beneath polUni in the MIT, holding all of that tenant's policies. Option C (Application Profile) is correct because an application profile (fvAp) is a managed object nested under a tenant in the MIT, grouping EPGs that share the same configuration. Option E (EPG) is correct because an endpoint group (fvAEPg) is a managed object contained within an application profile in the MIT, representing a collection of endpoints with common policy requirements. Option A (Interface) is not a distinct MIT object in this context, since physical interfaces are modeled as ports (e.g., l1PhysIf) rather than a generic 'Interface' component. Option D (VLAN) is not an MIT component; VLANs are represented as VLAN pools or encap values (fvnsVlanInstP) rather than a standalone MIT object named VLAN.

Answer

Interface

Answer

VLAN

Question 6

An engineer is implementing storage virtualization with thin provisioning on a Pure Storage array. Which three benefits does thin provisioning provide? (Choose three.)

Accepted Answer

Ability to over-allocate storage. Thin provisioning on a Pure Storage array lets the engineer present more logical capacity to hosts than physically exists, which is exactly the ability to over-allocate storage (A), since space is consumed only as writes occur. Because physical capacity can be purchased incrementally rather than sized for worst-case growth, it reduces upfront capacity cost (B). By avoiding stranded, pre-allocated but unused space, thin provisioning increases storage utilization (D), allowing the array to serve more workloads from the same physical footprint. Option C is incorrect because snapshot management is a separate data-protection feature (Pure's Purity snapshots and SafeMode) and is not a benefit of thin provisioning itself. Option E is incorrect because performance isolation is delivered through QoS mechanisms and volume/network design, not through thin provisioning, which concerns capacity allocation only.

Answer

Simplified snapshot management

Answer

Guaranteed performance isolation

Question 7

A UCS administrator needs to implement boot-from-SAN for a B-series blade. Which two components must be properly configured in the service profile? (Choose two.)

Accepted Answer

vHBA. Boot-from-SAN on a B-series blade requires the service profile to define the virtual Host Bus Adapter (vHBA), which provides the Fibre Channel connectivity and WWPNs used to reach the SAN fabric and target LUNs, so option A is correct. The service profile must also reference a Boot policy that specifies the SAN boot target (for example, the primary and secondary LUN/WWPN or boot order), so option C is correct. An IPMI policy (B) only governs out-of-band management access and has nothing to do with SAN boot. A SCSI policy (D) is not a valid service-profile component for this purpose; SAN boot uses vHBA and boot policy settings. A vNIC (E) provides Ethernet/LAN connectivity and is used for network boot scenarios, not Fibre Channel SAN boot.

Answer

IPMI policy

Answer

SCSI policy

Answer

vNIC

Question 8

Which three components are required for a basic Cisco UCS B-Series deployment? (Choose three.)

Accepted Answer

Blade Servers (B-Series). For a basic Cisco UCS B-Series deployment, the essential hardware components are the blade servers themselves, the Fabric Interconnects, and the blade chassis. Option B (Blade Servers (B-Series)) is correct because the B-Series blades are the compute nodes that host the workloads and are the defining element of a B-Series system. Option C (Fabric Interconnects) is correct because the Fabric Interconnects (e.g., UCS 6200 or 6300 series) provide the unified network and storage connectivity, management, and fabric switching that all B-Series components depend on. Option E (UCS 5108 Blade Chassis) is correct because the 5108 chassis physically houses and powers the B-Series blades and provides the backplane connections to the Fabric Interconnects via IOMs. Option A (UCS C-Series Rack Server) is not required because C-Series rack servers are a separate product line and are not part of a basic B-Series blade deployment. Option D (HyperFlex Cluster) is not required because HyperFlex is a hyperconverged infrastructure solution built on UCS servers, not a mandatory component of a basic B-Series deployment.

Answer

UCS C-Series Rack Server

Answer

HyperFlex Cluster

Question 9

An organization is deploying UCS Central to manage multiple UCS domains. Which THREE benefits does UCS Central provide over using individual UCS Managers?

Accepted Answer

Simplified firmware management across multiple domains.. Option B is correct because UCS Central provides centralized firmware repository and firmware management, allowing administrators to define firmware policies once and apply them consistently across all registered UCS domains, rather than updating each UCS Manager individually. Option C is correct because UCS Central aggregates inventory, faults, and health status from multiple UCS domains into a single unified dashboard, giving a single pane of glass for monitoring the entire environment. Option E is correct because UCS Central enables global service profiles and policies (such as global service profile templates, pools, and policies) that can be shared and deployed across domains, ensuring consistent configuration. Option A is not correct because automatic failover between Fabric Interconnects occurs within a single UCS domain via clustering, not across separate domains managed by UCS Central. Option D is not correct because direct management of blade chassis power supplies is handled by the individual UCS Manager within each domain, not by UCS Central.

Answer

Automatic failover between Fabric Interconnects in different domains.

Answer

Direct management of blade chassis power supplies.

Question 10

Which THREE factors should be considered when calculating the required number of buffer credits for a long-distance Fibre Channel link? (Choose three.)

Accepted Answer

The link data rate (e.g., 16 Gbps).. Option A is correct because the link data rate (e.g., 16 Gbps) determines how fast frames are transmitted, and buffer credits must be sufficient to keep the link full during the round-trip time at that speed. Option B is correct because the maximum frame size (e.g., 2148 bytes) affects how many frames can be in flight per credit, so larger frames require more buffer credits to sustain throughput. Option E is correct because the distance between the switches determines the round-trip propagation delay, which directly increases the number of outstanding frames that must be buffered. Options C and D are not correct because the number of VSANs and the number of zones in the fabric do not affect the physical link's buffer credit requirement; they are logical fabric constructs unrelated to the distance-delay product.

Answer

The number of VSANs configured.

Answer

The number of zones in the fabric.

Question 11

Which TWO statements correctly describe active zone sets in a Cisco MDS Fibre Channel fabric?

Accepted Answer

The active zone set is propagated to all switches in the fabric. Option A is correct because in a Cisco MDS Fibre Channel fabric, the active zone set is distributed fabric-wide via the zone server (using CFS distribution) so that every switch in the VSAN enforces the same zoning. Option B is correct because a VSAN can have only one active zone set at a time; activating a different zone set replaces the previously active one. Option C is incorrect because individual zones within an active zone set cannot be selectively deactivated — you must modify and reactivate the zone set. Option D is incorrect because the active zone set is stored in the zone database (nonvolatile) and is not limited to the running configuration. Option E is incorrect because only one zone set can be active per VSAN, not multiple simultaneously.

Answer

Zones within the active zone set can be selectively deactivated

Answer

The active zone set is stored in the running configuration only

Answer

Multiple zone sets can be active simultaneously

Question 12

A storage administrator is implementing synchronous replication between two storage arrays over a Fibre Channel link. Which three characteristics are true of synchronous replication? (Choose three.)

Accepted Answer

It requires a high-bandwidth, low-latency link between the arrays.. Option B is correct because synchronous replication must complete the write on both arrays before acknowledging the host, so the inter-array link needs sufficient bandwidth and very low latency (typically limited to distances where round-trip latency stays within the application's tolerance, often tens of kilometers over Fibre Channel). Option C is correct because in synchronous replication the host write I/O is only acknowledged after the data is committed to both the primary and secondary arrays, guaranteeing the secondary copy is identical at the moment of acknowledgment. Option D is correct because that dual-commit behavior means no data acknowledged to the host can be lost in a failover, yielding an RPO of essentially zero (the lowest possible). Option A is not correct because long-distance replication over hundreds of kilometers introduces latency that makes synchronous writes impractical; that scenario is served by asynchronous replication. Option E is not correct because synchronous replication generally consumes more bandwidth and is more latency-sensitive than asynchronous replication, which batches and sends writes without waiting for remote acknowledgment.

Answer

It is suitable for long-distance replication over hundreds of kilometers.

Answer

It uses less bandwidth than asynchronous replication.

Question 13

An engineer is deploying iSCSI storage in a data center. Which two features should be configured to improve performance and redundancy? (Choose two.)

Accepted Answer

Jumbo frames (MTU 9000). Jumbo frames (MTU 9000) (A) is correct because increasing the MTU on the iSCSI initiator, target, and every intervening switch port reduces per-packet header processing overhead and interrupts, raising throughput for large block I/O. MPIO (E) is correct because multipathing provides multiple redundant paths between initiator and target, enabling failover if a NIC, switch, or cable fails and allowing load balancing across paths for better performance. VSAN (B) is incorrect because it is a Fibre Channel fabric partitioning concept, not an iSCSI feature. CHAP authentication (C) is incorrect because it is a security mechanism for authenticating initiators, not a performance or redundancy feature. FCoE (D) is incorrect because Fibre Channel over Ethernet is a separate storage protocol that encapsulates FC frames, not an enhancement to an iSCSI deployment.

Answer

VSAN

Answer

CHAP authentication

Answer

FCoE

Question 14

Which TWO statements correctly describe the difference between synchronous and asynchronous replication?

Accepted Answer

Asynchronous replication has lower impact on application performance. Option B is correct because asynchronous replication acknowledges writes to the host as soon as they hit the local primary storage, without waiting for the remote site to confirm receipt, so the added write latency and throughput penalty on the application are minimal. Option C is correct because synchronous replication only returns write acknowledgment after the remote replica has committed the data, meaning every acknowledged write exists at both sites and the recovery point objective (RPO) is effectively zero. Option A is wrong because asynchronous replication does not require dedicated links; it is typically used over shared or higher-latency WAN links precisely because it tolerates distance and delay. Option D is wrong because synchronous replication is the least tolerant of high latency, since each write must complete a round trip to the remote site before acknowledgment. Option E is wrong because asynchronous replication can lose the most recent writes that have not yet been transmitted, so it cannot guarantee zero data loss.

Answer

Asynchronous replication requires dedicated links

Answer

Synchronous replication is more tolerant of high latency links

Answer

Asynchronous replication guarantees zero data loss

Question 15

A network administrator is planning a UCS C-Series rack server deployment. Which two statements correctly describe the integration modes with UCS Manager? (Choose two.)

Accepted Answer

In Direct Connect mode, the server must be connected to a Fabric Interconnect via a supported adapter.. Option C is correct because Direct Connect mode requires the C-Series server to be physically cabled to a Fabric Interconnect through a supported Cisco virtual interface card (VIC) adapter, which provides the data path for UCS Manager management. Option D is correct because in Direct Connect mode UCS Manager manages the rack server directly through the Fabric Interconnect, treating it as part of the UCS domain. Option A is incorrect because IMC Supervisor is a separate standalone management tool and does not integrate the server into UCS Manager. Option B is incorrect because in IMC Supervisor mode the CIMC is managed by IMC Supervisor, not directly by UCS Manager. Option E is incorrect because standalone mode means the server is managed by its own CIMC, not by UCS Manager, and it cannot be managed by UCS Manager without being integrated via Direct Connect or another supported mode.

Answer

In Cisco IMC Supervisor mode, the server is managed by the IMC Supervisor but can also be integrated into UCS Manager.

Answer

In Cisco IMC Supervisor mode, the server's CIMC is managed directly by UCS Manager.

Answer

Standalone mode allows the server to be managed by UCS Manager without any additional configuration.