Domain 4 Answer Key

For Endpoint Agent HTTP Server tests, valid alert rule metrics include Response Time and Error Type. BGP Reachability and Interface Throughput metrics are not applicable to this test type. Response Time measures time-to-first-byte while Error Type allows alerting on specific HTTP errors.

The correct answer is option A) Alert conditions weren't met and won't trigger with current setup because the condition requires 2 agents to generate an error.

Incorrect Options:

• B) Alert will trigger when the condition is met for a single test round
• C) Response code is correctly configured and should trigger for the HTTP code 401 shown in the exhibits
• D) Does not apply

The correct answer is D) Response Code is server error(5XX). This is the most specific and relevant condition for the given scenario.

Incorrect Options:

• A) HTTP Error: Too broad, capturing any error in the HTTP process (HTTP, Receive, Wait, SSL, etc.). While it includes HTTP server errors, it's not specific enough.
• B) Time to First Byte: Measures the duration between completing a browser request and receiving the first byte of the server's response. Not related to server error codes.
• C) Response Time: Measures overall request-response duration. Not directly related to HTTP response codes.

Option A) Scheduled tests, Hop# is correct because:

• Scheduled tests are suitable for monitoring consistent changes in the network path, like an ASN change on a specific hop. Real User Tests are better suited for monitoring real-time user experience fluctuations.
• Hop# is the appropriate condition as the requirement explicitly mentions monitoring a specific hop for the ASN change.

Incorrect Answers:

• B) Real User Test, Hop#: Real User Tests are not the best choice for this scenario as they are designed for monitoring real-time user experience, not network path changes.
• C) Scheduled tests, Any Hop: This option would trigger the alert if the ASN changes on any hop, not just a specific one.
• D) Real User Test, Path Length: This option would trigger the alert if the total number of hops in the path changes, not if the ASN changes on a specific hop.

Option C) Real User Tests > Endpoint, CPU utilization ≥ % is correct because:

• This combination directly targets the user's device (Endpoint) where the suspected CPU issue resides.
• It utilizes Real User Tests, which gather real-time performance data during user activity, providing the most accurate representation of the issue's impact.
• It specifically monitors for CPU utilization exceeding a defined threshold (≥ %), allowing for alerts to be triggered when CPU usage reaches problematic levels.

Observe the widgets on both dashboards to determine the test type.

The executive dashboard uses the map and number widgets.

The Mexico City Mexico (TelMex) Agent displays a Connect time of 0.51 ms.

The alert rule is displayed at the beginning of the dashboard, indicating the trigger reason.

Move your cursor over the map widget to the Mexico City agent to view the page completion time.

The number widget displays the total TE error count for the last 15 days.

Place your mouse over the latest metrics for page load time and DOM load time, then subtract the DOM load time from the page load time (1479.6 - 1354.3 = 125.3 ms).

The correct answer is:

• C) The alert rule's "Settings" section does not have the correct Web - HTTP Server test selected. This means the alert conditions are not being evaluated against the test's data, so the 500 errors are not triggering the alert.

Incorrect options:

• A) A disabled alert rule would prevent alerts, but this would be obvious in the Alert Rules page.

• B) If the test was not assigned to any agents, it would never generate data to trigger alerts. However, this would likely be noticed when viewing the test.

• D) A 200 OK response would not trigger the ">= 400" alert condition, but the question states 500 errors are occurring, so this is not the issue.

To make a CPU utilization Endpoint Agent alert less sensitive and reduce noise, the correct options are:

• A) Increasing the number/percentage of agents that must exceed the CPU threshold will prevent a single agent from triggering the alert.

• C) Requiring more rounds of data to be above the threshold (e.g. 2 of 3 rounds instead of 1 of 1) will filter out brief CPU spikes.

The incorrect options that would not reduce alert sensitivity are:

• B) Lowering the CPU utilization percentage would make the alert more sensitive and trigger more frequently.

• D) Enabling the alert on more agents would potentially trigger it more often, not less.

The correct answer is C) Tune the existing QoS configuration to prioritize voice and video traffic.

• This option directly addresses the observed issues (delay and jitter spikes) during peak hours.
• It aligns with the scope of the exam, which includes QoS tuning but not complete redesigns.
• This solution is targeted and likely more cost-effective than other options.

Incorrect options:

• A) A complete QoS redesign is out of scope for this exam and may be unnecessary.
• B) Increasing bandwidth on all links is a costly solution that may not specifically address the voice and video traffic issues.
• D) Replacing all network hardware is an extreme and costly solution that may not directly solve the problem.

The correct answer is B) Change the access point to a different, less congested channel.

• This directly addresses the high channel utilization issue.
• Reducing channel congestion can decrease retransmissions and improve overall voice call quality.
• This solution is a targeted optimization based on the SNMP data provided.

Incorrect options:

• A) Increasing transmit power may exacerbate interference issues and doesn't address channel congestion.
• C) Disabling all non-voice traffic is an extreme measure that could negatively impact other necessary network functions.
• D) Strict admission control for all clients doesn't specifically target the channel utilization issue and may be too restrictive.

The correct answer is B) Implement traffic shaping on non-voice traffic.

• This solution addresses the root cause by managing non-voice traffic to prevent it from overwhelming the voice queue.
• It's a targeted optimization that can reduce latency for voice traffic without major hardware changes.
• This approach aligns with QoS tuning, which is within the scope of the exam.

Incorrect options:

• A) Increasing queue size may delay packets further and doesn't address the underlying issue of queue saturation.
• C) Disabling QoS would likely worsen the situation for voice traffic, which requires prioritization.
• D) Replacing the router is an expensive solution that may not be necessary if the issue can be resolved through configuration changes.

This is a tricky question. Is there really an issue with the data being presented or perhaps there is something misconfigured on the platform?

Our exhibit shows that our highest consumption, although marked at 97%, is merely 48Mbps, certainly not enough to be making use of the entire 1Gbps connection from the ISP, so option A would be incorrect. Even though our top traffic is indeed HTTP, nothing in the exhibit indicates that pruning some specific HTTP traffic could fix how data is being presented, so option C is incorrect. The exhibit also fails to provide any reason as to how changing business hours could provide some benefit in this case, so option D is incorrect.

Finally, if we gather all the data we have available: ISP connection is 1Gbps and capacity planning marks 48Mbps as 97% of max capacity, we can reach the conclusion that the max capacity for this interface is misconfigured; it should be set to 1Gbps instead of the value it currently has, thus, option B is correct.