Skip to main content
Practice set 1 is free for every certification. Unlock sets 2-5 for $10.
Practice Exam

300-445 ENNA — Cisco Certified Specialist - Enterprise Network Assurance

Test your knowledge with official exam-style questions

Questions25PassingPass/Fail (scaled score not published)Exam time90 min

Questions and options are shuffled each attempt

Email me this practice set

Get a link sent to your inbox so you can pick this back up anytime.

▶Cisco Certified Specialist - Enterprise Network Assurance — Practice Set 1: All Questions & Explanations

Full question text, answer options, and explanations for this practice set — a spoiler-free alternative is the interactive quiz above for scored, shuffled practice.

  1. 1. A network engineer reads RFC 7799 to classify a monitoring approach that injects dedicated test packets into the network in order to measure delay. Which statement correctly describes an Active Method of measurement?

    • A. It is based solely on observations of an undisturbed and unmodified packet stream of interest
    • B. It is a combination of Passive and Active Methods of measurement
    • C. It generates packet streams for measurement, usually with fields dedicated to measurement, and can influence the quantities it measures to some degree(correct)
    • D. It relies on observation points exporting their observations to a collector without any added traffic

    Explanation: RFC 7799 states that Active Methods generate packet streams, that the packets usually carry fields dedicated to measurement, and that adding traffic to the network means Active Methods influence the quantities measured to some degree. Option A is incorrect: That is the RFC 7799 description of a Passive Method, which observes an undisturbed and unmodified stream. Option B is incorrect: That is the definition of a Hybrid Method, which uses a combination of Active and Passive Methods. Option D is incorrect: Exporting observations to a collector without adding traffic is an aspect of Passive Methods, not Active Methods.

  2. 2. A company wants a ThousandEyes vantage point that it deploys itself inside its own network, behind its firewalls, and that is dedicated to its organization rather than shared with other customers. Which agent type fits this requirement?

    • A. Cloud Agent
    • B. Endpoint Agent
    • C. Enterprise Agent(correct)
    • D. Cloud Agent hosted at an internet exchange point

    Explanation: Enterprise Agents are Linux-based software agents that you deploy and manage within your own network infrastructure (data centers, branch offices, IaaS), and they are dedicated to your organization and not shared with other customers. Option A is incorrect: Cloud Agents are managed and maintained by the ThousandEyes Operations team and are available to all ThousandEyes customers, so they are not dedicated to one organization or deployed behind its firewalls. Option B is incorrect: Endpoint Agents run natively on Windows and macOS end-user machines; they are not the dedicated, Linux-based agents that an organization deploys inside its infrastructure. Option D is incorrect: Cloud Agents located in internet exchange points are still ThousandEyes-managed, shared vantage points.

  3. 3. A branch office has only Microsoft Hyper-V hosts and no Linux servers. The team wants to run a ThousandEyes Enterprise Agent at the branch. Which approach is supported?

    • A. Install a native Enterprise Agent installer on a Windows Server virtual machine
    • B. Deploy the ThousandEyes Virtual Appliance (TEVA) on Hyper-V(correct)
    • C. Run the Enterprise Agent Docker container on Docker for Windows
    • D. Install the Enterprise Agent Linux package directly on the Hyper-V host operating system

    Explanation: The TEVA is a ready-to-use virtual machine that includes the operating system and agent software and can be deployed on VMware ESXi, Oracle VirtualBox or Microsoft Hyper-V. ThousandEyes states there is no native Enterprise Agent for Windows or macOS and that a TEVA hosted in a virtualization product must be used. Option A is incorrect: There is currently no native version of the Enterprise Agent software for Microsoft Windows or macOS. Option C is incorrect: Docker running on Windows and Mac is not currently supported for Enterprise Agents. Option D is incorrect: The Linux package is installed on a supported Linux operating system such as Ubuntu or Red Hat, not on the Windows-based Hyper-V host.

  4. 4. Which statement best describes how ThousandEyes WAN Insights produces its recommendations for a Cisco SD-WAN?

    • A. It analyzes vAnalytics telemetry of application flow records from SD-WAN routers and recommends alternate paths at the site level for each application category, and the network team then applies the change(correct)
    • B. It runs synthetic HTTP tests from Endpoint Agents and automatically pushes new application-aware routing policy into vManage
    • C. It collects BGP updates from public monitors and withdraws affected prefixes on WAN edge routers
    • D. It polls WAN edge interfaces by SNMP and generates recommendations for each individual user

    Explanation: WAN Insights uses vAnalytics network telemetry, analyzing application data flow records from the routers in the SD-WAN fabric, and generates path recommendations for application categories defined in vManage. Recommendations are made at the site level per application category, and the path changes are made outside WAN Insights by the network team, for example through a localized SD-WAN policy. Option B is incorrect: WAN Insights is not based on Endpoint Agent synthetic tests, and it does not push policy automatically; the required action is to reconfigure the network equipment yourself. Option C is incorrect: WAN Insights does not collect BGP updates or modify prefixes on WAN edge routers. Option D is incorrect: WAN Insights relies on vAnalytics telemetry rather than SNMP polling, and its recommendations apply to sites and application categories, not individual users.

  5. 5. An operations team wants ThousandEyes alerts to create incidents automatically in their ITSM platform, ServiceNow, so that existing incident workflows handle them. Which integration approach matches this requirement?

    • A. Stream ThousandEyes metrics, logs and traces in OpenTelemetry format to a Grafana instance
    • B. Install ThousandEyes Enterprise Agents on Catalyst switches using Catalyst Center
    • C. Enable the Meraki data enrichment connector so Meraki LAN data appears in path visualization
    • D. Configure the ServiceNow alert notification integration so ThousandEyes alert notifications are routed into ServiceNow incident management(correct)

    Explanation: ThousandEyes supports routing alert notifications directly into ServiceNow, delivering them into the ServiceNow account to be processed by that system's workflows, with incident management as one of the supported methods. Option A is incorrect: OpenTelemetry streaming exports telemetry to observability platforms; it is not the mechanism that routes alert notifications into ServiceNow incidents. Option B is incorrect: The Catalyst Center integration installs Enterprise Agents on Cisco devices; it does not route alerts to an ITSM. Option C is incorrect: Meraki data enrichment adds Meraki LAN data to path and segment visualizations; it does not create incidents.

  6. 6. A test engineer must configure a ThousandEyes HTTP Server test against an internal API that requires a bearer token. Which Scheme option in the test's HTTP authentication settings is intended for this?

    • A. Basic
    • B. Kerberos
    • C. OAuth(correct)
    • D. Verify SSL Certificate

    Explanation: The HTTP Server test Scheme setting offers Basic, Kerberos and OAuth. OAuth uses a bearer token for modern API authentication. Option A is incorrect: Basic uses a simple username and password rather than a bearer token. Option B is incorrect: Kerberos uses network credentials for single sign-on in enterprise environments. Option D is incorrect: Verify SSL Certificate validates the server certificate against the agent trust store; it is not an authentication scheme for sending a token.

  7. 7. Which transport protocol is NOT supported for ThousandEyes agent-to-server network measurements but IS supported by agent-to-agent tests?

    • A. UDP(correct)
    • B. ICMP
    • C. TCP
    • D. TCP using SACK-based measurement

    Explanation: ThousandEyes states that while UDP is not supported by the agent-to-server network measurements (UDP has no protocol-level facility that makes a target respond), it is supported by agent-to-agent tests, where agents at both ends exchange the probe stream. Option B is incorrect: ICMP echo request and reply is one of the three protocols (TCP, UDP, ICMP) and is supported for agent-to-server measurements. Option C is incorrect: TCP is supported for agent-to-server measurements, using TCP SYN probes and expecting SYN+ACK or RST responses. Option D is incorrect: SACK-based TCP measurement is a TCP variant that ThousandEyes TCP-based network tests prefer, so it is supported.

  8. 8. An administrator will deploy the ThousandEyes Endpoint Agent MSI to Windows laptops by using Group Policy. Select the TWO requirements that ThousandEyes lists for this deployment method.

    • A. A shared network drive from which the Endpoint Agent installer is accessible(correct)
    • B. The ability to create Group Policy Objects in Active Directory(correct)
    • C. A Munki web server with munkiimport configured
    • D. The Microsoft Win32 Content Prep Tool to convert the MSI
    • E. A Docker host running the Enterprise Agent container

    Explanation: The Group Policy method lists three requirements: a shared network drive, the Endpoint Agent being accessible via that network drive, and the ability to create Group Policy Objects in Active Directory. Option C is incorrect: Munki and munkiimport are prerequisites of the macOS deployment method, not Group Policy on Windows. Option D is incorrect: The Win32 Content Prep Tool is used to convert MSI packages for Intune deployment, not Group Policy. Option E is incorrect: A Docker host is unrelated; the Endpoint Agent is installed natively on Windows end-user devices.

  9. 9. A DNS administrator configures a ThousandEyes DNS Server test to query an authoritative name server directly and wants that server to answer from its own zone data rather than recurse. How should the Send recursive queries option be set?

    • A. Enabled, so that the agent sets the Recursion Desired (RD) flag
    • B. Enabled, because it is mandatory for any test that queries port 53
    • C. Enabled, because it is required whenever the record type is A
    • D. Left disabled, which is the default(correct)

    Explanation: The Send recursive queries option sets the Recursion Desired flag, is available only when Class is IN, and is disabled by default. ThousandEyes advises leaving it disabled when you want to query an authoritative server directly. Option A is incorrect: Enabling the option sets the RD flag, which requests recursive processing; that is the opposite of the stated goal. Option B is incorrect: The option is not mandatory; it is disabled by default and every agent queries each configured server on port 53 regardless. Option C is incorrect: The record type (A, AAAA, MX and so on) is a separate setting and does not require recursion.

  10. 10. When creating a ThousandEyes RTP Stream test to simulate a voice call, which DSCP value is applied to the test packets by default?

    • A. AF41 (DSCP 34)
    • B. EF (DSCP 46)(correct)
    • C. CS3 (DSCP 24)
    • D. Default forwarding (DSCP 0)

    Explanation: The RTP Stream test sets the DSCP value in the IP header of test packets to simulate a QoS policy, and it defaults to EF (DSCP 46) for high-priority voice traffic. Option A is incorrect: AF41 is a valid DSCP marking but not the RTP Stream test default. Option C is incorrect: CS3 is a valid DSCP marking but not the RTP Stream test default. Option D is incorrect: DSCP 0 would give the stream no priority marking, whereas the default is EF to mimic voice traffic.

  11. 11. Which of the following targets cannot be used as the URL of a ThousandEyes Page Load test?

    • A. https://portal.example.com/index.html, an HTML page
    • B. The URL of an XHTML application
    • C. https://cdn.example.com/banner.jpg, an image file(correct)
    • D. https://intranet.example.com/, which returns an HTML document

    Explanation: Page load tests only work for HTML pages or XHTML applications, so the URL cannot point to an image such as a .jpg or .png file. Option A is incorrect: An HTML page is exactly the content type page load tests are designed for. Option B is incorrect: XHTML applications are explicitly supported by page load tests. Option D is incorrect: A URL that returns an HTML document is a valid page load target.

  12. 12. In a bidirectional ThousandEyes agent-to-agent test, one-way packet loss is 20% from source to target and 10% from target to source. What does the Both Directions loss metric report?

    • A. 15%
    • B. 30%
    • C. 28%(correct)
    • D. 10%

    Explanation: Both Directions loss is the combined probability that a packet is lost on the round trip: 100 - ((100 - 20) x (100 - 10) / 100) = 100 - 72 = 28%. Option A is incorrect: 15% is the simple average, but the platform states Both Directions loss is not a simple average of the two one-way values. Option B is incorrect: 30% would be the sum of the two percentages; the loss streams are independent and combine as probabilities. Option D is incorrect: 10% only reflects the target-to-source direction and ignores the loss in the other direction.

  13. 13. In Path Visualization for a web server test, a red ring appears on the rightmost destination node, and no intermediate hop shows forwarding loss. What does this indicate?

    • A. Terminal loss: packets reach the target's network but the target itself is dropping them, so the cause is likely the application, firewall, load balancer or host(correct)
    • B. Forwarding loss inside an upstream ISP between the agent and the target
    • C. Loss confined to the agent's own local network
    • D. A BGP route leak affecting the destination prefix

    Explanation: ThousandEyes distinguishes forwarding loss (a red ring mid-path, where packets drop in transit) from terminal loss (a red ring on the destination node, where packets reach the target's network but the target drops them). Terminal loss points at or near the target: application, firewall, load balancer or target host. Option B is incorrect: A mid-path red ring would indicate forwarding loss in the agent's local network, an ISP, a peering exchange or transit; this scenario shows no intermediate loss. Option C is incorrect: Loss in the agent's local network would appear as a ring on early hops, not on the destination node. Option D is incorrect: Path visualization loss rings do not identify route leaks; those are routing events seen in BGP monitoring.

  14. 14. In a ThousandEyes waterfall chart for a page load test, which component measures the time between the browser finishing sending a request and receiving the first byte of the server's response?

    • A. BLOCKED
    • B. CONNECT
    • C. RECEIVE
    • D. WAIT(correct)

    Explanation: WAIT is the duration between the completion of the browser's SEND and receipt of the first byte of the server's response, so a high WAIT points at server processing or time to first byte. Option A is incorrect: BLOCKED is the time a browser waits for an already established connection to become available. Option B is incorrect: CONNECT is the time to establish the TCP handshake with the target server. Option C is incorrect: RECEIVE is the time between the first byte of the response and the last byte of the data payload.

  15. 15. Remote users on a corporate VPN show heavy packet loss in ICMP-based Endpoint Agent scheduled tests, yet their applications work normally. The VPN is known to block ICMP. What does ThousandEyes recommend?

    • A. Reduce the test interval to one minute
    • B. Use TCP-based testing(correct)
    • C. Enable VPN support in the Endpoint Agent settings
    • D. Replace the Endpoint Agent with an Enterprise Agent on each laptop

    Explanation: ThousandEyes recommends TCP-based testing because some VPNs block ICMP traffic. It also notes that large enterprise networks often block ICMP entirely. Option A is incorrect: Changing the interval does not change which protocol the probes use, so ICMP would still be blocked. Option C is incorrect: No additional configuration is required to enable VPN support; it works once the agent detects a supported VPN. Option D is incorrect: Enterprise Agents are Linux-based and are not installed on end-user laptops.

  16. 16. A ThousandEyes DNS Trace test for a company domain fails with the error 'responded with AA flag unset (non-authoritative answer)'. What does this error indicate?

    • A. The DNSSEC chain of trust is broken and the response is Bogus
    • B. The final authoritative name server returned SERVFAIL
    • C. A server in the path returned an answer without the Authoritative Answer flag set, so the trace cannot terminate; spoofed or otherwise non-authoritative responders can cause this(correct)
    • D. The agent's HTTP request to the web server timed out

    Explanation: A DNS Trace terminates successfully only when the Answer section has non-CNAME records and the AA flag is set. Non-authoritative nameservers, spoofed or otherwise, can populate the Answer section without AA, and in that case the trace returns this different error message, which is a lead worth investigating for DNS spoofing. Option A is incorrect: Bogus or Invalid validity is a DNSSEC test result, not the cause of this specific DNS Trace error. Option B is incorrect: A SERVFAIL or NXDOMAIN from the final server is a different failure; this error is specifically about a response lacking the AA flag. Option D is incorrect: HTTP timeouts belong to web layer tests and are unrelated to a DNS Trace error.

  17. 17. A BGP test for a company's /24 prefix shows 100% reachability from all monitors, but users report that the hosted application cannot be reached. How should the engineer interpret the BGP data?

    • A. The application server must be up because the prefix is reachable
    • B. Packet delivery is confirmed because routes exist for the prefix
    • C. Reachability proves that ICMP works end to end from every monitor
    • D. It only means monitors have an active route to the prefix; it does not verify packet delivery or application availability, so other tests are needed(correct)

    Explanation: ThousandEyes states that BGP reachability means a monitor has an active route to a prefix and does not verify packet delivery or application availability. BGP monitoring shows routing information rather than traffic behavior. Option A is incorrect: Route presence says nothing about whether the server is running. Option B is incorrect: BGP monitoring provides routing visibility, not packet delivery confirmation. Option C is incorrect: BGP monitors do not send ICMP; they analyze route advertisements.

  18. 18. Why would ThousandEyes Internet Insights not report an application outage for an application that only one customer uses?

    • A. Because Internet Insights only analyzes DNS results
    • B. Because application outages also require data from multiple customers' tests, and an outage is not reported for an application used by a single customer(correct)
    • C. Because application outages are detected only for internet service providers
    • D. Because Internet Insights only examines BGP updates

    Explanation: Internet Insights draws on de-identified data from the collective set of tests run on the platform. Application outages require data from multiple customers' tests, so one customer's failing application is not reported as an outage event. Option A is incorrect: Internet Insights analyzes network-layer and application-layer test results, not only DNS. Option C is incorrect: Application outages are displayed by affected domain; ISPs are a separate provider type in the catalog. Option D is incorrect: Internet Insights examines path visualization and test results from Cloud and Enterprise Agent tests, not only BGP.

  19. 19. Which set of periodic local network measurements does the ThousandEyes Endpoint Agent collect every minute?

    • A. Ping and path trace to the system gateway, and a DNS trace to the configured resolver
    • B. A browser HAR capture and DNS resolution time for every visited domain
    • C. Ping and path trace data to detected proxies and VPNs, ping data to the system gateway, and a network profile(correct)
    • D. A 100-packet train to the destination and a bandwidth measurement

    Explanation: Every minute the Endpoint Agent collects ping and path trace data to the proxies and VPNs currently detected, ping data to the system's gateway, and a network profile. Local probes use automatic protocol detection to choose ICMP or TCP. Option A is incorrect: The gateway receives ping probes, not a path trace, and a DNS trace is not one of the periodic local probes. Option B is incorrect: HAR capture is part of browser-based data collection triggered by site visits, not the periodic local probes. Option D is incorrect: A 100-packet train and bandwidth measurement belong to Enterprise Agent bandwidth measurements, not the Endpoint Agent's local probes.

  20. 20. An administrator creates a custom alert rule by clicking Add New Alert Rule and does not change the severity field. Which severity does the rule receive?

    • A. Info(correct)
    • B. Minor
    • C. Major
    • D. Critical

    Explanation: A customized rule created with Add New Alert Rule is automatically configured with the severity Info. Rules based on ThousandEyes default alert rules are configured with Minor, and either can be changed. Option B is incorrect: Minor is the severity given to alert rules created from Default Alert Rules, not custom ones. Option C is incorrect: Major is selectable but is not an automatic default. Option D is incorrect: Critical is selectable but is not an automatic default.

  21. 21. An alert rule is set so that the same agent must exceed a threshold in three consecutive rounds. The Atlanta agent trips it in round 1, the Ashburn agent in round 2, and the San Francisco agent in round 3. What happens?

    • A. An alert triggers because three different agents tripped the rule
    • B. No alert triggers, because 'the same' requires the same agent to meet the condition in consecutive rounds(correct)
    • C. An alert triggers for San Francisco only
    • D. An alert triggers after the next round regardless of which agent trips it

    Explanation: Selecting 'the same' creates sticky triggers: the alert activates only if the same set of alert triggers meets the condition in consecutive rounds. Here a different agent tripped in each round, so no alert activates. 'Any of' would behave differently. Option A is incorrect: Three different agents is exactly the situation 'any of' would count; with 'the same' it does not satisfy the rule. Option C is incorrect: No single agent met the condition in three consecutive rounds, so no alert is raised for San Francisco. Option D is incorrect: A further round does not change that the same agent never tripped three rounds in a row.

  22. 22. A ThousandEyes quantile dynamic baseline uses the formula Alert threshold = (Qupper - Qlower) x k + Qlower + 1. The upper quantile (Q90) is 100 ms, the lower quantile (Q10) is 10 ms and the threshold factor k is 2. At what value will the alert fire?

    • A. 190 ms
    • B. 110 ms
    • C. 200 ms
    • D. 191 ms(correct)

    Explanation: Applying the formula: (100 - 10) x 2 + 10 + 1 = 191 ms, so the alert fires for any value above 191 ms. Option A is incorrect: 190 ms omits the +1 term in the formula. Option B is incorrect: 110 ms is Q90 + Q10 and does not apply the threshold factor. Option C is incorrect: 200 ms is 2 x Q90 and ignores the lower quantile and the +1 term.

  23. 23. An HTTP Server test shows an average response time of 500 ms over the last 24 hours. An alert rule uses a classic dynamic baseline with the percentage-change method at its default setting. Above which response time will the next test round fire the alert?

    • A. 520 ms
    • B. 550 ms
    • C. 600 ms(correct)
    • D. 572 ms

    Explanation: The percentage-change method defaults to 20% above the baseline, so the alert fires above 500 ms + 20% = 600 ms. Option A is incorrect: 520 ms would be a 4% change, not the 20% default. Option B is incorrect: 550 ms corresponds to the absolute-value method with a 50 ms default for HTTP tests, not percentage change. Option D is incorrect: 572 ms is the result of a standard-deviation example (500 + 36 x 2), not the percentage-change method.

  24. 24. A network team schedules planned maintenance on a data center link and applies a scheduled Alert Suppression Window to the affected Enterprise Agent tests. Which statement describes the effect?

    • A. Data from the assigned tests is not evaluated against alert rule criteria during the window, so alerts cannot trigger until it ends, and notifications do not occur(correct)
    • B. Alerts still trigger and are listed as active, but notifications are queued and sent when the window ends
    • C. The tests stop collecting data for the duration of the window
    • D. All alerts in the account are suppressed, including agent offline notifications

    Explanation: For scheduled suppression windows, data from assigned tests is not evaluated against alert rule criteria, so data meeting the conditions does not trigger an alert until the window ends; notifications that would normally occur do not occur during the window. Option B is incorrect: Notifications are not queued; they do not occur during the window, and the data is not evaluated for alerting. Option C is incorrect: The window affects alert evaluation and notifications, not test data collection. Option D is incorrect: Only tests or alerts assigned to the window are affected; notifications for events such as Enterprise Agents going offline or online are not.

  25. 25. An engineer needs to give a third-party ISP a point-in-time view of a ThousandEyes dashboard without giving them a platform login. What should the engineer use?

    • A. Share the dashboard directly with the ISP's email address
    • B. Deploy a dashboard template for the ISP
    • C. Open an Internet Insights outage dashboard for the ISP
    • D. Create a dashboard snapshot and turn link sharing on(correct)

    Explanation: Dashboard snapshots capture a specific time range and can be shared with people outside your organization when link sharing is on, for example a third-party provider or ISP. Sharing a dashboard directly is possible only within your organization, with ThousandEyes users. Option A is incorrect: Direct dashboard sharing is limited to other ThousandEyes users in your organization. Option B is incorrect: Dashboard templates help you create dashboards quickly and consistently; they are not a way to share data externally. Option C is incorrect: Internet Insights dashboards show outage data from the collective intelligence; they are not a mechanism for sharing your own dashboard.