N10-009 domain - 14% of the exam

Network Security

Network Security is 14% of the N10-009 exam. These are the objectives it covers, each with practice questions, with every answer explained.

The domain in numbers

42
Practice questions
3
Objectives
42
Medium

Where people go wrong

  • Matching a control to the wrong CIA principle

    The wrong answers pick a real control that serves a different CIA principle. HTTPS protects confidentiality in transit, yet a downloaded firmware image still needs a SHA-256 comparison to prove integrity. Backups serve availability, so they cannot stop a thief reading a stolen laptop, where full-disk encryption protects confidentiality. A message authentication code mismatch reveals alteration, which makes it an integrity check even though the code uses a secret key.

    See it in a sample question
  • Strengthening an existing layer instead of adding one

    The wrong answers reinforce a control that already exists. A faster firewall, a hot-standby firewall, a second packet-filtering firewall, a stronger edge firewall and stronger passwords each deepen or duplicate one layer. Defence in depth needs a new, independent control elsewhere. Host-based firewalls and antivirus sit behind the perimeter, VLANs with inter-VLAN access control lists contain lateral movement inside the LAN, and a web application firewall inspects HTTP requests that a network firewall permitting port 443 cannot see into.

    See it in a sample question
  • Matching a denial-of-service symptom to the wrong attack

    Each wrong answer names a genuine attack that exhausts a different resource. A full backlog of half-open connections in SYN_RECEIVED shows a TCP SYN flood; amplification would instead saturate bandwidth with large UDP responses. Small spoofed UDP queries that make resolvers send far larger responses to a victim show DNS amplification, and no TCP handshake is left half-open. A nearly idle link with connections trickling partial HTTP headers shows Slowloris, which ties up worker threads at the application layer.

    See it in a sample question
  • Confusing ARP poisoning with other redirection attacks

    The wrong answers swap ARP poisoning for another attack that also redirects or disrupts traffic. MAC flooding overflows the switch CAM table, DHCP starvation exhausts leases, and DNS poisoning corrupts hostname-to-IP answers. ARP poisoning sends forged replies that bind the gateway's IP address to the attacker's MAC address in victims' caches within the local segment. A forged A record held in a recursive resolver until the TTL expires points to DNS cache poisoning, because ARP works at Layer 2 inside one subnet.

    See it in a sample question
  • Blaming direction or protocol for ACL processing

    The wrong answers reach for interface direction, a trailing permit or a protocol rule when the cause is how ACLs are processed. Entries are matched top-down and stop at the first match, so a subnet permit placed above a host deny means the host is never tested against the deny. Every ACL ends with an implicit deny, which drops DNS once only TCP 80 and 443 are permitted. A standard ACL matches the source address alone, so one destination port needs an extended ACL.

    See it in a sample question

Objectives in this domain

What to study

The official documents this domain's practice questions cite most.

Sample questions from this domain

Free sampleNetwork Securitymedium

Before flashing a downloaded switch firmware image, an engineer wants to confirm that the file was not altered or corrupted in transit. Which action most directly protects the integrity of the firmware?

  • ACompute the image's SHA-256 hash and compare it with the vendor-published checksum before installing it. Correct
  • BDownload the image again over an HTTPS link so that the whole transfer is encrypted end to end.
  • CStore the image on a RAID-protected server that is backed up to tape every night.
  • DRequire multi-factor authentication before the engineer may reach the vendor download portal.
Verifying a downloaded file against a published cryptographic hash protects integrity by proving the contents were not altered or corrupted during transfer. Integrity is verified by comparing a recomputed cryptographic hash such as SHA-256 with the vendor-published value; a match confirms the file's contents are unchanged, while encryption or redundancy address different principles.

Why A is correct: A cryptographic hash uniquely fingerprints the file; matching the computed SHA-256 value against the vendor checksum proves the contents were not altered or corrupted, which is exactly integrity.

Why B is wrong: HTTPS encrypts the transfer and protects confidentiality in transit, but it does not prove the resulting file matches the vendor original, so it does not verify integrity.

Why C is wrong: RAID and backups keep a copy available if hardware fails, which serves availability; they do nothing to confirm the downloaded contents were not modified.

Why D is wrong: MFA strengthens who can authenticate to the portal, but controlling access to the download is not the same as verifying that the file's contents remain unchanged.

Free sampleNetwork Securitymedium

A public web application is currently protected only by a network firewall that permits inbound TCP port 443. To follow defence in depth, an engineer wants to add a control that operates at a different layer to protect the application itself. Which addition meets that goal?

  • AAdd a second network packet-filtering firewall in series that permits inbound TCP port 443 to the web server, matching the first device.
  • BReconfigure the web application to listen on a non-standard high port instead of 443 so that it is harder for attackers to find.
  • CEnable an account-lockout policy on the firewall's management console after several failed administrator login attempts.
  • DDeploy a web application firewall that inspects HTTP requests for SQL injection and cross-site scripting before they reach the application. Correct
Defence in depth adds controls at different layers, so a web application firewall protects the application layer where a network firewall cannot inspect HTTP content. A network firewall filters by port and address but cannot inspect HTTP content; a web application firewall adds an application-layer control that examines requests for injection and cross-site scripting, forming a genuinely different defensive layer.

Why A is wrong: A second packet-filtering firewall permitting the same port operates at the same network layer as the first, adding redundancy but no protection against attacks carried inside legitimate HTTPS traffic.

Why B is wrong: Moving the service to an unusual port is security by obscurity, and it neither inspects application traffic nor adds a real defensive layer, since the port is easily discovered by scanning.

Why C is wrong: Locking the firewall's management console protects administrative access to the device, but it does nothing to defend the web application against malicious requests reaching it over port 443.

Why D is correct: A web application firewall inspects the HTTP payload for injection and cross-site scripting, adding an application-layer control that the network firewall cannot provide, which is a genuine additional layer.

Free sampleNetwork Securitymedium

A public web server becomes unreachable during a sudden traffic spike. On the server, netstat shows thousands of half-open connections stuck in the SYN_RECEIVED state, each from a different spoofed source address, and the connection backlog is full so legitimate clients time out. Which attack best explains this behaviour?

  • AA TCP SYN flood, in which the attacker sends a stream of SYN segments without completing the three-way handshake, filling the backlog with half-open connections. Correct
  • BAn ARP poisoning attack, in which the attacker binds false MAC addresses to the server's IP so its traffic can be intercepted on the local segment.
  • CA DNS amplification attack, in which small spoofed queries elicit much larger responses that overwhelm the inbound bandwidth of the target link.
  • DA VLAN hopping attack, in which double-tagged frames let the attacker reach the server's VLAN and exhaust the resources of hosts on it.
A TCP SYN flood exhausts a server's connection backlog with spoofed half-open handshakes, denying service to legitimate clients. The TCP three-way handshake reserves connection state as soon as a SYN arrives; by sending many spoofed SYNs and never sending the final ACK, the attacker keeps connections in SYN_RECEIVED until the backlog is full and further clients are turned away.

Why A is correct: Each spoofed SYN forces the server to reserve state and reply with SYN-ACK while the final ACK never arrives, so the backlog of half-open connections fills and new clients are refused.

Why B is wrong: ARP poisoning is a real attack, but it is an on-path interception technique operating at Layer 2, not a method that fills a TCP connection backlog with half-open handshakes.

Why C is wrong: Amplification is also a flooding attack, but it saturates bandwidth with large UDP responses and does not create half-open TCP states in a SYN_RECEIVED backlog.

Why D is wrong: VLAN hopping is a genuine Layer 2 attack, but it is about crossing VLAN boundaries with tagged frames, not about filling a server's TCP connection backlog.

Free sampleNetwork Securitymedium

On a switched LAN, by what mechanism does ARP poisoning let an attacker insert itself onto the path between a host and its default gateway?

  • AIt overflows the switch's MAC address table so that frames are flooded across the switch's ports, forcing traffic to pass the attacker.
  • BIt alters DNS responses so that the gateway's hostname resolves to an IP address that the attacker controls on the segment.
  • CIt sends unsolicited ARP replies that bind the gateway's IP address to the attacker's MAC address in the victims' ARP caches. Correct
  • DIt exhausts the DHCP address pool so that clients are forced to accept the attacker as their new default gateway.
ARP poisoning forges ARP replies to map the gateway's IP to the attacker's MAC, redirecting a victim's traffic through the attacker on the local segment. ARP includes no authentication, so an attacker can broadcast forged ARP replies that overwrite the victim's cache, associating the gateway IP with the attacker's MAC and steering the victim's frames to the attacker first.

Why A is wrong: This describes MAC flooding, a genuine Layer 2 attack, but it indiscriminately floods traffic rather than selectively binding the gateway address to the attacker, which is how ARP poisoning redirects a specific conversation.

Why B is wrong: This confuses ARP poisoning with DNS poisoning; it is tempting because both redirect traffic, but ARP operates on IP-to-MAC bindings within the local segment, not on hostname resolution.

Why C is correct: ARP has no authentication, so forged gratuitous ARP replies overwrite a victim's cache entry for the gateway, causing frames destined for the gateway to be sent to the attacker's MAC address instead.

Why D is wrong: This mixes ARP poisoning with DHCP starvation and a rogue server; it is plausible as a gateway hijack, but ARP poisoning corrupts existing IP-to-MAC cache entries rather than manipulating address leases.

Free sampleNetwork Securitymedium

An inbound ACL on a router interface lists, in order: line 10 permit ip 192.168.10.0 0.0.0.255 any, then line 20 deny ip host 192.168.10.50 any. The technician intended line 20 to block host 192.168.10.50, but that host still passes traffic. What is the most effective fix?

  • AAdd an explicit permit ip any any statement to the end of the ACL so the deny for the host will then take effect.
  • BChange line 20 to a standard ACL entry, because extended entries cannot deny traffic from a single host address.
  • CReorder the entries so the deny for host 192.168.10.50 sits above the permit for the 192.168.10.0/24 subnet, because the ACL stops at the first match. Correct
  • DApply the ACL in the outbound direction instead of inbound so that the deny statement is evaluated after routing.
ACLs are processed top-down and stop at the first match, so a broad permit placed above a specific deny lets the targeted traffic through. Router ACLs evaluate entries sequentially and act on the first match. Because line 10 permits the whole 192.168.10.0/24 subnet, host 192.168.10.50 matches and is permitted before line 20 is reached, so the specific deny must sit above the subnet permit.

Why A is wrong: It is tempting because ACLs do end with an implicit deny, but a trailing permit does not change that line 10 already matches the host first, so 192.168.10.50 stays allowed.

Why B is wrong: Extended ACL entries can match a single host with the host keyword, so converting to a standard entry is unnecessary and would lose destination matching; the real problem is statement order.

Why C is correct: ACLs are processed top-down and stop at the first match; with the subnet permit first, the host matches line 10 and is never tested against line 20, so moving the host deny above it blocks the host.

Why D is wrong: Changing direction alters where packets are filtered relative to routing but not the first-match order within the list, so the subnet permit would still match the host before the deny.

Other domains in this exam

See also the N10-009 cert hub, the study guide, and the cheat sheet.

Examworthy is not affiliated with or endorsed by CompTIA. Original, blueprint-aligned practice material only.