CCNP ENARSI 300-410 practice questions, every option explained
260 exam-style questions for the advanced routing concentration exam of CCNP Enterprise, with timed exam simulation and readiness per domain. Try five below.
3 days and 100 questions, no card needed.
Try 5 questions, no account needed
Answer each one to see why it is right or wrong.
-
Question 1 of 5Layer 3 Technologies
R1 and R2 are joined by a GigabitEthernet link in OSPF area 0. R1's interface has ip ospf network point-to-point, while R2's keeps the default broadcast type. Hello and dead intervals are 10 and 40 seconds on both, and nothing else differs. What is the result?
- The adjacency stops at 2-WAY, because R1 takes no part in the DR election
- No adjacency forms, because the hellos carry the network type and it differs
- The routers reach FULL, but SPF cannot use the link, so no routes cross it
- The routers reach FULL and route normally, because only the timers must match
Show the answer
Answer: C. The routers reach FULL, but SPF cannot use the link, so no routes cross it. Each router describes the link differently in its router LSA: R1 as a point-to-point link to R2, R2 as a transit link to a DR network. SPF requires both ends to describe the same link (the two-way check), so the databases synchronize but no routes through the link are installed. OSPF never compares network types in the hello, so a point-to-point/broadcast mismatch with equal timers still reaches FULL. The damage shows afterwards: the two router LSAs describe the link in incompatible ways and SPF ignores it. When neighbors are FULL but routes are missing, compare show ip ospf interface on both ends.
-
Question 2 of 5VPN Technologies
PE2 must classify traffic arriving from the core on the MPLS TC (EXP) bits, so the engineer adds mpls ldp explicit-null on PE2. P2 is the penultimate hop for PE2's loopback. What changes for packets that are label-switched to PE2's loopback?
- PE2 advertises label 0 for its loopback, so P2 swaps the top label to 0 and PE2 still reads the TC bits
- PE2 advertises label 3 for its loopback, so P2 swaps the top label to 3 and PE2 reads the TC bits on arrival
- PE2 advertises label 0 for its loopback, so P2 removes the whole label stack and PE2 classifies on IP DSCP
- PE2 advertises label 0 to its neighbors, so every LSR on the path swaps the top label to 0 until PE2
Show the answer
Answer: A. PE2 advertises label 0 for its loopback, so P2 swaps the top label to 0 and PE2 still reads the TC bits. Explicit null (label 0 for IPv4) replaces the implicit null PE2 advertised before. P2 now swaps the top label to 0 instead of popping it, so the packet reaches PE2 still labeled, with the TC bits set in the core; PE2 then removes label 0 and continues with the next label or the IP lookup. Label 3 (implicit null) is a signal, never a label on the wire: it tells the penultimate hop to pop. Label 0 (explicit null for IPv4, label 2 for IPv6), advertised after mpls ldp explicit-null, makes the penultimate hop swap to 0 instead, so the egress LER still receives a labeled packet and can use its TC bits before it pops label 0.
-
Question 3 of 5Infrastructure Security
Two routers authenticate administrator logins against the same ISE node: R4 uses TACACS+ and R5 uses RADIUS with PAP. An analyst has a full packet capture of both exchanges but does not know either shared secret. What can the analyst read?
- Both usernames, since each protocol hides only the password field
- The RADIUS username, but nothing inside the TACACS+ packet bodies
- The TACACS+ username, but not the one in the RADIUS Access-Request
- Neither username, since both protocols encrypt everything after the header
Show the answer
Answer: B. The RADIUS username, but nothing inside the TACACS+ packet bodies. RADIUS protects only selected attributes such as User-Password, so User-Name, the NAS address and the other attributes are in clear text. TACACS+ XORs the entire body with an MD5-based pad derived from the shared secret; only its 12-byte header (version, type, sequence number, flags, session ID, length) stays readable. TACACS+ (TCP 49) protects the whole payload, so usernames, authorized commands and accounting records are hidden from a capture. RADIUS (UDP 1812/1813) hides only the password and a few other attributes. That, plus TACACS+'s separate handling of authentication, authorization and accounting, is why device administration usually runs on TACACS+ and network access on RADIUS.
-
Question 4 of 5Infrastructure Services
A new collector accepts only IPFIX. R1 exports Flexible NetFlow as NetFlow v9 today, and you change its flow exporter to export-protocol ipfix. Which statement is true?
- You must also rewrite the flow record, because v9 match and collect fields cannot be sent as IPFIX
- IPFIX has no templates; it uses a fixed record format like NetFlow v5, so the collector needs no setup
- Export packets now carry version number 10 and still describe the data with templates, as v9 does
- IPFIX runs only over TCP, so the collector must accept a TCP session from R1 before export starts
Show the answer
Answer: C. Export packets now carry version number 10 and still describe the data with templates, as v9 does. IPFIX was standardized from NetFlow v9: its message header carries version 10, and data sets are still described by template sets that the exporter sends to the collector. NetFlow v5 has a fixed, IPv4-only format; v9 (RFC 3954) introduced templates so that any set of fields can be exported; IPFIX (RFC 7011), with version number 10, standardizes v9 and adds variable-length and enterprise fields. In Flexible NetFlow the record stays the same and only the exporter's export-protocol changes.
-
Question 5 of 5Layer 3 Technologies
R1's show ip bgp summary lists a neighbor in the OpenConfirm state. What has happened on that session so far?
- TCP is up, R1 has sent its own OPEN, and it is still waiting for the peer's OPEN
- R1 has sent a TCP SYN to port 179 and is waiting for the connection to complete
- TCP is up, R1 accepted the peer's OPEN, sent a KEEPALIVE and waits for one
- R1's connection attempt failed, so it listens for the peer while it retries
Show the answer
Answer: C. TCP is up, R1 accepted the peer's OPEN, sent a KEEPALIVE and waits for one. In OpenConfirm the OPEN exchange is done: R1 checked the peer's OPEN (AS, BGP identifier, hold time) and sent a KEEPALIVE. The peer's KEEPALIVE moves it to Established; a NOTIFICATION or hold-timer expiry sends it back to Idle. BGP states in order: Idle (not trying, or blocked), Connect (TCP being opened), Active (TCP attempt failed, retrying and listening), OpenSent (own OPEN sent, waiting for the peer's), OpenConfirm (OPENs exchanged, waiting for a KEEPALIVE) and Established. A neighbor stuck in Connect or Active has a TCP problem; one that reaches OpenSent and falls back has an OPEN problem, such as a wrong AS.
Exam domains covered
Weights follow Cisco’s published exam topics for ENARSI 300-410 v1.1. Most of the exam asks you to troubleshoot, not just configure.
- Layer 3 Technologies35%Administrative distance, route maps, redistribution, summarization, policy-based routing, VRF-Lite, BFD, EIGRP, OSPFv2 and v3, BGP
- VPN Technologies20%MPLS operations and Layer 3 VPN, DMVPN with a single hub: GRE and mGRE, NHRP, IPsec, spoke-to-spoke
- Infrastructure Security20%IOS AAA with TACACS+ and RADIUS, IPv4 ACLs, IPv6 traffic filters, uRPF, CoPP, IPv6 first hop security
- Infrastructure Services25%Device management, SNMP, logging and conditional debugs, DHCP for IPv4 and IPv6, IP SLA, NetFlow, DNA Center assurance
Inside the question bank
260 questions you can practise, 260 of them with a written explanation.
| Domain | Questions | Answered correctly |
|---|---|---|
| Layer 3 Technologies | 87 | — |
| Infrastructure Services | 65 | — |
| VPN Technologies | 64 | — |
| Infrastructure Security | 44 | — |
From the CertAce database, November 2026. A rate is shown once a domain has at least 150 answers.
Why every option is explained
I failed my first CCNA attempt with 58%. I had memorised answers without understanding why the other options were wrong, and the exam found every gap. CertAce is the tool I wished I had then: for each question you see why the right answer is right and why each wrong one is wrong.
Adrian, network engineer and founder of CertAce. Read the full story.
Questions people ask
How long is the ENARSI 300-410 exam?
90 minutes. Cisco does not publish a fixed number of questions; candidates usually report around 55 to 65, so plan on about a minute and a half per question. ENARSI is a concentration exam: together with ENCOR it gives you the CCNP Enterprise, and on its own it earns the Cisco Certified Specialist, Enterprise Advanced Infrastructure Implementation.
Should I take ENCOR or ENARSI first?
Cisco does not set an order. Most people take ENCOR first, because ENARSI assumes you already know OSPF, EIGRP and BGP and then asks you to troubleshoot them: redistribution, route maps, summarization and path selection. Taking ENARSI soon after ENCOR, while the routing is fresh, saves a lot of revision.
What score do I need to pass ENARSI?
Cisco does not publish the passing score, and the score report is scaled. Book the exam when you pass several full timed quizzes here comfortably, in every domain, not after your first good one.
Which ENARSI domains are hardest?
Layer 3 Technologies is 35% of the exam and most of it is troubleshooting from show output: a redistribution loop, a route map that filters more than intended, a BGP attribute that changes the best path. DMVPN and MPLS Layer 3 VPN catch out people who have never configured them.
Is CertAce free?
Yes, to start. The five questions on this page need no account. The free trial gives you 3 days and 100 questions from the full bank, no card needed, and the plans are on the pricing page.
Read next
CertAce is not affiliated with or endorsed by Cisco Systems. CCNP and Cisco are trademarks of Cisco Systems, Inc.