An application is hosted on Google Kubernetes Engine. A new JavaScript module is created to work with the existing application. Which task is mandatory to make the code ready to deploy?

Create a Dockerfile for the code base.
Rewrite the code in Python.
Build a wrapper for the code to “containerize” it.
Rebase the code from the upstream git repo.

Answer(s): A

Show Answer Next Question

DRAG DROP (Drag and Drop is not supported)
Drag and drop the steps from the left into the correct sequence on the right to describe how to use Git to maintain the current HEAD and revert back to a previous commit, while undoing all intermediate commits.

Select and Place:

See Explanation section for answer.

Answer(s): A

Explanation:

Show Answer Next Question

QUESTION: 13

Which database type should be used with highly structured data and provides support for ACID transactions?

time series
document
graph
relational

Answer(s): D

Show Answer Next Question

QUESTION: 14

Where should distributed load balancing occur in a horizontally scalable architecture?

firewall-side/policy load balancing
network-side/central load balancing
service-side/remote load balancing
client-side/local load balancing

Answer(s): D

Show Answer Next Question

QUESTION: 15

Which two statements about a stateless application are true? (Choose two.)

Different requests can be processed by different servers.
Requests are based only on information relayed with each request.
Information about earlier requests must be kept and must be accessible.
The same server must be used to process all requests that are linked to the same state.
No state information can be shared across servers.

Answer(s): A,B

Show Answer Next Question

Download PDF Next Page

Viewing page 4 of 78
Viewing questions 16 - 20 out of 507 questions

Post your Comments and Discuss Cisco 350-901 exam prep with other Community members:

Comments:

Name:

350-901 Exam Discussions & Posts
350-901 AI Verifications

Alex on: March 02, 2025
Great material
UNITED KINGDOM

WhosJohnAcronym on: May 11, 2026
Real exm questions made this exam seem challenging when it was just very hard on its own. I got through by relying on brain dumps but barely passed.
Italy

AzureNinja on: May 23, 2026
It took a while but the practice questions helped me understand what I needed. I spent more time reviewing than I thought necessary. Passed eventually with relief after all that effort.
Pakistan

finnk8s on: July 25, 2026
After long workdays, I could only squeeze in short study sessions for this exam. It was a relief when the practice questions on this website helped me finally get through it.
Spain

weisyseng on: June 08, 2026
The exam was super stressful. Thankfully the practice questions here made it manageable.
Italy

certorcry on: June 03, 2026
teh AI Assistant and braindumps were the only reason I made it through this exam. Very hard and I wasn't sure I'd pass.
Kenya

anyacloudarch on: June 03, 2026
This exam was tough but the prep questions on this site really got me through.
Australia

deskjobdreamer on: June 19, 2026
This exam definitely covered more than I thought it would so felt kinda overwhelming. The AI assistant on this site was pretty helpful though in keeping me from getting lost in all the info.
India

dmitrilinuxpro on: July 10, 2026
That challenging exam drained me but the exam dumps made a difference. The stress level was high yet those real exam quetions were crucial.
Ukraine

dockerdave on: June 20, 2026
At first I wasn’t feeling super confident about this exam but the practice questions on this website really helped me learn. Going over them again and again made a big difference in understanding what I needed to focus on.
Saudi Arabia

hamidcertguy on: July 14, 2026
I wasn't feeling too confident initially but the practice questions really helped me learn a lot. This website made things much clearer and gave me a better understanding of what to expect on the exam.
Ireland

n0d3n on: April 22, 2026
Passed it with the help of braindumps and the AI Assistant but it was a very hard exam. It took a toll on my schedule but the real exam questions prepped me well.
Japan

PowerShellPete on: May 23, 2026
I was not confident at first but going through the practice questions here really helped to understand what might come in this exam.
Ireland

LastMinuteLearner on: July 01, 2026
Initially underestimated this exm but it was very hard so I relied heavily on brain dumps and real exam questions to get through.
United States

mateusawsbr on: July 22, 2026
Balancing work and study was tough so I did short bursts of practice questions whenever I could. Felt a huge relief when it was finally over and all the prep from this website really helped me pass.
Portugal

PowerShellPete on: July 04, 2026
Juggling work and this exam prep was tough but manageable. The practice questions here really helped me pass in the end.
Brazil

chiaradevsec on: July 14, 2026
I felt like this exam covered way more than I was ready for. The prep material from the website was still pretty useful so that helped a bit.
Poland

sith on: July 17, 2026
Can anyone confrim, if these are the actual exam questions ?
Ancaster, Canada

chiaradevsec on: May 07, 2026
I've been juggling work and short study sessions with practice questions. Finally done with the exam so that's a weight off my shoulders.
United States

tomcertmaster on: June 09, 2026
The real exam questions were very hard and I only got through it by using brain dumps. Definitely felt the stress throughout.
Egypt

OracleCertV on: July 17, 2026
Juggling work and short study sessions was rough, but I felt a huge relief after finally getting through this exam. This website really helped simplify things for me with its practice questions.
Australia

rachelops on: July 22, 2026
I squeezed in exam prep between projects at work whenever I had a free moment. It took time, but this website helped me pass the exam eventually.
South Korea

gitadataeng on: June 10, 2026
Barely scraped through this exam after weeks of stress only because those brain dumps were spot on. Real exam questions were a nightmare to handle.
Israel

JuniperJeff on: June 02, 2026
This exam was harder than I thought and covered way more topics. The material here helped a lot and the practice questions were pretty spot on.
Norway

t1_support_guy on: June 05, 2026
This exam was very hard and I barely made it through after using brain dumps. The stress was real but those brain dumps kept me from completely losing it.
Japan

alwayslearninga on: July 08, 2026
I'm juggling a lot at work but managed to squeeze in some study time here and there with the help of this website. It took me a while but I passed the exam on my first try.
Pakistan

PassOrFailLol on: May 02, 2026
The exam was much more challenging than I anticipated and I really relied on the brain dumps to finally pass it. Without the AI Assistant I doubt I would have gotten through those real exam questions so fast.
Singapore

NetwrkKing on: April 30, 2026
Finally done with this exam after endless hours spent on brain dumps and the AI Assistant. Not the easiest path but the real exam quetions matched and saved me.
Brazil

finnk8s on: June 29, 2026
At first, I didn't feel confident about this exam at all but the practice questions helped me a lot. I learned quite a bit just by working through relevant questions over and over again on this website.
Spain

laylaituae on: July 22, 2026
Studying for this exam felt like it took forever. The practice questions helped focus my review time though. I finally passed so worth it in the end.
UAE

5igmas on: June 09, 2026
The exam was very hard and I wasn't sure I would pass but the AI Assistant and braindumps pulled me through. Real exam questions felt familiar after using the dumps for practice.
Israel

TechNerd92 on: May 28, 2026
At first I not feeling so confident but going through practice questions helped me understand more for this exam.
Philippines

gitadataeng on: July 04, 2026
Honestly the exam was a bit tougher than I thought but manageable thanks to practice questions here.
Germany

flashcardfanatic on: July 07, 2026
Finally done with this challenging exam using brain dumps and teh AI Assistant. Honestly thought I might not pass but those dumps were a relief.
United States

CiscoFanJ on: May 29, 2026
I’ve been squeezing in study sessions between meetings and client calls so progress was slow. Thankfully the practice questions on this site were spot on which helped me pass the exam.
Saudi Arabia

overtimeomar on: April 29, 2026
The exam dumps helped with the challenging exam but it was still very hard to get through.
South Africa

justanothertechguy on: June 15, 2026
The exam was tough but these questions were super relevant and saved me.
China

omaritpro on: June 13, 2026
Took two attempts to pass this exam using brain dumps after struggling hard to understand the materials. Real exam questions can be very hard without the extra help.
Turkey

miguelcloudops on: June 01, 2026
Between work and life finding time to study was rough. Short study sessions using practice questions from this site really paid off in the end and I'm so relieved it's done.
Bangladesh

4wscert on: July 09, 2026
Three weeks of prep with braindumps and the AI Assistant finally paid off despite doubting I’d make it through this challenging exam. Real exam questions matched well but it was still very hard.
Rwanda

CertHunter on: July 07, 2026
Juggling work and cramming for this exam with short study sessions was pretty exhausting. Finally getting through it felt like a huge relief thanks to the relevant questions on this website.
Kenya

dantheadmin on: July 15, 2026
That exam was seriously stressful. The prep material here definitely helped a lot.
Japan

liamsecops on: April 28, 2026
This exam was very hard adn I underestimated it so I had to grind through many brain dumps to finally pass.
Kenya

miket2024 on: July 08, 2026
At first I didn't feel very confident but the practice questions on this website helped me a lot.
Thailand

vlanjockey on: June 16, 2026
This was a very hard exam and the brain dumps helped a lot to fill gaps in my study. Spent a lot of late nights reviewing the dumps to pull through and I am relieved it's done.
Bangladesh

miguelcloudops on: June 03, 2026
My progress was slow at first but going through the practice questions over and over really helped. It took some time to get it right but finally passed this exam with decent score.
India

ExamWarrior on: August 08, 2026
The exam had a lot more depth than I was expecting which threw me off. That AI assistant on this site really came through for me though with some handy pointers.
Mexico

5igmas on: July 02, 2026
Took two attempts to clear this exam adn only managed with the help of exam dumps and some real exam questions. The AI Assistant didn't help that much with a very hard set of questions.
Colombia

sshhero on: July 03, 2026
The exm dumps didn't cover everything so the real exam questions were surprisingly tough. Felt pretty unprepared despite weeks of studying the braindumps.
United Kingdom

CoffeeAndCerts on: May 17, 2026
Initially I felt like my progress was pretty slow with the prep. But taking extra time to review really helped me grasp everything better. Finally passed this exam after a few attempts.
South Africa

AI Explanation - verified by Poornachander1995 on: July 30, 2026
Question 18:
Question 18 asks when to deploy the Splunk_TA_ForIndexes.spl add-on to indexers in an ES deployment.

Correct answer: C) After installing ES on the search head(s) and running the distributed configuration management tool.

Why:

ES requires the Splunk_TA_ForIndexes.spl add-on on indexers to apply index-time configurations. It should be deployed via the distributed configuration management (DCM) flow after ES is installed on the search heads, so the indexers receive the proper configuration in a controlled, versioned bundle.

Why the other options are not correct:

A) Deploying when adding apps to the deployment server isn’t the ES-specific order; it bypasses the ES-instantiated workflow.

B) Installing Splunk_TA_ForIndexes.spl first isn’t the recommended order; ES needs to be installed on the search heads first.

D) It isn’t limited to indexer cluster sites or cluster master commands; the standard flow uses DCM to push the TA to indexers in general.

Takeaway: install ES on the search heads, run DCM to push configurations, then deploy Splunk_TA_ForIndexes.spl to indexers to complete the ES indexing setup.
Bengaluru, India

AI Explanation - verified by User9198 on: July 27, 2026
Question 30:
Question 30 asks which two are pillars of the AWS Well-Architected Framework.
Correct answers: B and E (Reliability and Operational Excellence).
Why:

The Well-Architected Framework has five pillars: Operational Excellence, Security, Reliability, Performance Efficiency, and Cost Optimization.

Operational Excellence: how you run and improve your systems—monitoring, incident response, and evolving processes.

Reliability: the ability of a workload to recover from failures, maintain uptime, and scale as demand changes.

Why the others aren’t pillars:

Availability (A) is not a pillar name; it’s an outcome that relates to Reliability.

Scalability (C) isn’t a pillar name; scalability touches on Performance Efficiency.

Responsive design (D) is not a pillar in the Well-Architected Framework.

Busan, South Korea

AI Explanation - verified by User2505 on: August 10, 2026
Question 50:
Correct answer: A — Purchase Provisioned Throughput for the custom model.
A model customized or fine-tuned in Amazon Bedrock cannot generally be invoked using the standard on-demand capacity used by foundation models. The company must purchase Provisioned Throughput, which allocates dedicated inference capacity for the custom model.
Why the other options are incorrect:

SageMaker endpoint: Not required because the model is being used through Amazon Bedrock.

SageMaker Model Registry: Used for managing SageMaker models, not for making a Bedrock custom model available.

Grant access in Amazon Bedrock: IAM permissions may still be needed, but granting access alone does not provide the required inference capacity.

London, United Kingdom

AI Explanation - verified by User7324 on: August 12, 2026
Question 56:
Correct answer: B — Recommend that the product owner reviews the backlog refinement processes.
Unclear requirements usually indicate that the product backlog is not being sufficiently refined. The product owner should work with stakeholders and the team to clarify acceptance criteria, split large items, prioritize work, and ensure backlog items are “ready” before iteration planning.
Why the others are weaker:

Review iteration length: Changing duration does not resolve unclear requirements.

Limit work in progress: Useful when excessive multitasking is the problem, but it does not directly clarify requirements.

Review individual performance: The issue is a process and backlog-quality problem, not necessarily an individual performance problem.

The key principle is: Improve backlog refinement and shared understanding before the team commits to iteration work.
Ho Chi Minh City, Vietnam

AI Explanation - verified by User5015 on: July 22, 2026
Question 10:
Question 10 asks if migrating all VMs to an Azure pay-as-you-go subscription uses the “elastic expenditure model.”
Here’s the right way to think about it:

In Azure, the typical cost approach for pay-as-you-go VMs is an Operational Expenditure (OPEX) model. You’re billed for actual usage each month.

The term “elastic expenditure model” isn’t a standard Azure cost category. Elasticity describes scaling capacity, not the billing category itself.

So recommending the elastic expenditure model is not correct for this goal.

Conclusion:

The solution does not meet the goal. Answer: No (B).

Why this matters:

Pay-as-you-go aligns with OPEX because costs vary with usage.

Elasticity is about scaling resources, which affects performance and capacity, not the fundamental cost model.

If you want, I can compare this with the other questions in this set (11 and 12) to highlight how the correct expenditure models differ in each scenario.
San Antonio, United States

AI Explanation - verified by rashmica.ts.9 on: July 19, 2026
Question 10:
Correct answer: E. “Tasks run in parallel via slots.”
Why:

In Spark, an executor is a JVM that can have multiple cores. Each task can consume some CPUs (spark.task.cpus).

The executor’s resources are effectively partitioned into slots, where each slot can run one task. The number of slots per executor is roughly spark.executor.cores / spark.task.cpus.

Therefore, multiple tasks can run in parallel within the same executor (via these slots). So the statement “Tasks run in parallel via slots” is correct.

Why the other options are wrong:

A: A slot is not another name for an executor; a slot is a unit inside an executor.

B: There’s no requirement that there must be fewer executors than tasks; tasks can be scheduled over time across the available slots.

C: An executor can run more than one core; it’s not limited to a single core.

D: There isn’t a rule that there must be more slots than tasks; tasks can wait if slots aren’t available, and you can have equal or even fewer slots than concurrent tasks.

Bengaluru, India

AI Explanation - verified by enbore2015 on: August 07, 2026
Question 1:
Correct answer: A — The BCP has not been tested since it was first issued.
A business continuity plan is only useful if it is current, workable, and proven through testing. Without testing, the organization cannot know whether:

Recovery procedures actually work

Recovery time and recovery point objectives can be met

Staff understand their responsibilities

Contact details, dependencies, and technical configurations are accurate

Gaps or conflicts exist in the plan

The other issues are important but generally less severe:

Not version-controlled: Creates a risk of using an outdated copy, but the plan might still be effective.

Contact information needs updating: This is a specific weakness that can be corrected, often identified during testing.

Not approved by senior management: Approval is important for authority and resource commitment, but an untested plan presents the most direct risk that recovery will fail when needed.

The exam principle is: testing provides assurance that the BCP is effective, not merely documented.
Nairobi, Kenya

AI Explanation - verified by User6701 on: August 10, 2026
Question 1:
Correct answer: A — Configure SetupComplete.cmd in %windir%\setup\scripts.
SetupComplete.cmd is a Windows Setup mechanism that runs automatically after Windows installation or deployment finishes. It can call your PowerShell scripts, for example:

cmd powershell.exe -ExecutionPolicy Bypass -File C:\Scripts\Configure-VM.ps1

This is appropriate when preparing a VM image or deploying multiple VMs that require the same post-installation configuration.
Why the other options are less suitable:

GPO logon script: Runs only when a user logs on, so configuration may be delayed and is user-dependent.

GPO startup script: Can work, but requires the VMs to be domain-joined and introduces GPO dependency. It is not specifically tied to Windows deployment.

Place scripts in a VHD: Merely makes the files available; it does not execute them.

Important correction: The supplied explanation discusses Azure Custom Script Extension, which does not match this question’s on-premises Hyper-V scenario. The answer key’s A is reasonable, but that explanation appears to belong to a different Azure VM question.
Amsterdam, The Netherlands

AI Explanation - verified by User3436 on: July 20, 2026
Question 1:
Correct answer: a, b, and c
Explanation:

a) True — To add a field to the form layout, drag the field from the Fields tab to the desired destination on the form. This adds to the form view only, without changing the table definition.

b) True — To create a new field on the form’s table, drag a data type from the Field Types tab to the form and configure it. This creates a new column on the table and adds the field to the form view.

c) True — To remove a field from the form layout, hover to enable the action buttons and select Delete. This removes the field from the form view but does not remove it from the table definition.

d) False — You cannot add a section by dragging from the Field Types tab. Sections are added via the Add Section option (toolbar or right-click menu), not from the Field Types tab.

Mumbai, India

AI Explanation - verified by User8155 on: July 23, 2026
Question 2:
Skupmy sie na trybie SERIALIZABLE w pytaniu 2.

Co to znaczy: transakcje zachowuja sie tak, jakby byly wykonywane jeden po drugim. Dzieki temu nie ma brudnych odczytów, niepowtarzalnych odczytów ani phantom reads.

Implementacja: PostgreSQL uzywa SSI (Serializable Snapshot Isolation). W praktyce oznacza to, ze jesli wykryje konflikt miedzy transakcjami, jedna z nich zostanie zakonczona bledem “serialization failure” i trzeba ja ponowic.

Przyklad (phantom read i serializowalnosc):

Zalózmy tabele orders(customer_id int, amount int).

Transakcja T1:

- BEGIN; - SELECT COUNT(*) FROM orders WHERE customer_id = 42; - -> powiedzmy zwraca 5.

Transakcja T2, równoczesnie:

- BEGIN; - INSERT INTO orders (customer_id, amount) VALUES (42, 100); - COMMIT;

T1 ponownie wykonuje to zapytanie:

- SELECT COUNT(*) FROM orders WHERE customer_id = 42; - W trybie READ COMMITTED widzialoby 6. - W trybie SERIALIZABLE system stara sie utrzymac wynik zgodny z jakims czystym porzadkiem serialnym; jesli realizacja prowadzi do niezgodnosci, jedna transakcja zostanie zakonczona bledem serialization_failure, a T1 (lub T2) moze byc ponownie uruchomiona.
Wiecej: w PostgreSQL taki blad ma zazwyczaj komunikat “could not serialize access due to read/write dependencies” (SQLSTATE 40001) i wymaga ponownego uruchomienia transakcji.
Warsaw, Poland

AI Explanation - verified by elisamangano02 on: August 03, 2026
Question 51:
Question 51 shows: let x = ('15' + 10) * 2; What is the value of a? (There’s likely a typo: there’s no a variable; the code assigns to x.)
Reasoning:

'15' is a string. With + and a string, JavaScript does string concatenation: '15' + 10 -> '1510'.

The result '1510' is then used in a multiplication: '1510' 2. The operator triggers numeric conversion, so '1510' becomes the number 1510.

1510 * 2 = 3020.

So, x equals 3020. The correct option is A.
Note:

If the code were let x = (15 + 10) * 2, the result would be 50 since 15 + 10 is numeric addition.

Sassari, Italy

AI Explanation - verified by User2505 on: August 10, 2026
Question 11:
Correct answer: A — Amazon SageMaker Feature Store
Amazon SageMaker Feature Store is a centralized repository for storing, organizing, and sharing ML features—the variables used by models, such as customer age, purchase frequency, or average order value.
It supports:

Reusing features across different models and teams

Consistent feature definitions

Managing feature versions and metadata

Using features for both training and inference

Why the others are incorrect:

SageMaker Data Wrangler: Prepares and transforms data; it is not primarily a shared feature repository.

SageMaker Clarify: Helps detect bias and explain model predictions.

SageMaker Model Cards: Documents models, including intended use, risks, and evaluation results.

The key phrase is “share and manage variables for model development”—these variables are ML features, making Feature Store the appropriate choice.
London, United Kingdom

AI Explanation - verified by User2940 on: July 20, 2026
A? ??? ?? ??
A? ??? ?? ??? ?? ??? ?? ?????.

A? ??? Aurora Replicas? ?? ??? ???? ?? ???, ?? ???? Network Load Balancer(NLB)? ??????. NLB? ?? 4 ?? ???? HTTP ??? ?? ???(Sticky sessions)? ????? ???? ????.

??? ?? ??????? ?? ?? ???? ??? ???? ?? ????? ????? ?? ?? ???(?? ?/??? ??)? ?????. ? ??? ALB?? ?? ????, NLB?? ???? ?????.

??? ??? ??? ??? ???? ?? ???? ???? HTTP ???? ??? ALB? ???? ???. A? ? ??? ??????.

??: ?????? ?? ??? ????, ?? ???? ???? ?? ??? ??? ???? ??? ????? ???? ????.
Seongnam-Si, South Korea

AI Explanation - verified by User4053 on: August 10, 2026
Question 52:
Correct answer: C — AWS Security Token Service (AWS STS).
AWS STS provides temporary security credentials—an access key, secret key, and session token—with:

A limited lifetime

Permissions defined by an IAM role or policy

The ability to access AWS APIs without using long-term IAM user credentials

An application can use STS operations such as AssumeRole to obtain credentials for a specific, restricted task.
Why the others are incorrect:

Amazon API Gateway — Builds and manages APIs; it does not generally issue AWS credentials for applications.

IAM users — Usually use long-term credentials, which are less suitable for applications.

IAM instance profiles — Provide temporary role credentials to applications running on EC2 instances, but are specifically tied to EC2. The broader requirement described is best met by AWS STS.

Itapira, Brazil

AI Explanation - verified by User6509 on: July 28, 2026
and why we can wipe option ?
Short answer: You could use a wipe, but it’s not the right choice for redeploying to a new user with Autopilot.

What wipe does: A factory reset that erases all data, apps, and settings and returns the device to the OOBE. It removes the device from management and resets identity.

Why not for this scenario: The goal is to repurpose the same device to a new user with Autopilot. Autopilot Reset keeps the device’s Azure AD/Intune identity and re-provisions the device for the next user without a full wipe.

What Autopilot Reset does: Removes user data and apps but preserves device identity and management connection, then reapplies the Autopilot provisioning for the next user.

When you’d use wipe instead: If you’re selling or permanently discarding the device, or you truly want to scrub all traces and start fresh without re-enrolling.

So, for Question 34’s redeployment flow, Autopilot Reset (not Wipe) is the correct action to take after the CSV upload. The wipe option exists for complete disposal, not for quick redeployment.
Lisbon, Portugal

AI Explanation - verified by User1537 on: July 26, 2026
Question 7:
Question 7 asks: What describes how components and activities work together to facilitate value creation?

Correct answer: A: The ITIL service value system (SVS).

Why A is correct:

The SVS explains how all components (governance, the service value chain, practices, guiding principles, information and technology, partners and resources, and continual improvement) work together to enable value creation from demand to outcomes.

Why the others are not correct:

B: The ITIL guiding principles describe aspirational guidance for decision-making, not how all components/activities interact.

C: The four dimensions of service management explain areas to consider (organizations/people, information/technology, partners/suppliers, value streams/processes) but don’t describe the integrated system itself.

D: A service relationship defines the relationship between provider and consumer, not the overall system of components and activities creating value.

Louisa, United States

AI Explanation - verified by User4569 on: July 31, 2026
Question 7:
Question 7 asks how to confirm integrity of a hashed message.

Correct answer: D — use the same hashing algorithm as the sender's to create a numerical representation of the file.

Why:

Integrity is verified by recomputing the hash on the received data and comparing it to the sender’s hash. This must use the same hash function; if you use a different algorithm, the digests won’t match even if the data is intact.

The hash output is a fixed-length digest (a numerical representation). Whether you call it a “binary image” or a “numerical representation,” the key is that you generate the same type of digest with the same algorithm and compare them.

Incorrect options:

A: same algorithm but the result is described as a binary image instead of a digest; you still compare the digest, but the phrasing is less precise.

B/C: using a different algorithm would produce a different digest, causing false mismatch or failure to verify integrity.

In short: recompute the hash with the same algorithm and compare the resulting digest to the one sent.
Sarajevo, Bosnia And Herzegovina

AI Explanation - verified by User5149 on: August 07, 2026
Question 5:
The answer key shows D and E, but this question appears ambiguous.

Permanent table (E): Correct. A dropped permanent table can be restored with UNDROP TABLE during its Time Travel retention period. Permanent tables also have Fail-safe protection.

Transient table: It can also generally be restored during its shorter Time Travel period, although it has no Fail-safe. Therefore, excluding it makes the question questionable.

Temporary table: Usually cannot be recovered after it is dropped; its lifecycle is limited to the session.

Materialized view (D): The answer key treats it as recoverable, but recovery behavior depends on the specific Snowflake object/version capabilities. It should not be confused with an ordinary view.

For exam purposes, the intended pair is likely D and E, emphasizing objects with Time Travel/Fail-safe protection. However, remember the key distinction: permanent objects have the strongest recovery protection; transient objects have limited Time Travel but no Fail-safe.
Pune, India

AI Explanation - verified by User9493 on: July 22, 2026
Question 36:
Correct answer: IDS (intrusion detection system)
Why:

An IDS monitors network traffic for signs of attacks and raises alerts. It can be deployed out-of-band (via a SPAN/mirror port or a tap) so it analyzes traffic without being inline, meaning it doesn’t measurably affect latency.

An IPS (intrusion prevention system) is inline and can block threats, but that processing can introduce some latency and potential throughput impact.

A firewall also inspects traffic and can add latency, though usually modest with proper hardware, but its primary purpose isn’t continuous passive analysis.

A load balancer distributes traffic and isn’t used for intrusion analysis.

So, to analyze attacks from the internet without impacting latency, use an out-of-band IDS.
Essex, United States

AI Explanation - verified by User8744 on: August 12, 2026
Question 11:
Question 11: Correct answer — C
> Set expectations and agree on the requirements that the contractor will deliver with periodic progress reports.
The project manager should delegate the work clearly while retaining appropriate oversight. This means:

Define the contractor’s deliverables, responsibilities, acceptance criteria, and deadlines.

Agree on a reporting cadence to monitor progress.

Allow the local company to use its expertise and manage its own work.

This empowers the contractor without abandoning control of the project.
Why the others are weaker:

A: “Under the supervision of the project manager” suggests excessive direct control and is less empowering.

B: Micromanaging prevents the contractor from functioning independently and is inefficient.

D: Deploying the contractor’s staff under the PM’s supervision creates a staffing/control arrangement rather than empowering the contracted company.

The exam principle is delegate outcomes and monitor performance, rather than personally directing every task.
Johor Bahru, Malaysia

AI Explanation - verified by User7799 on: July 20, 2026
Question 18:
Answer: D

What D means: In Security Command Center (SCC), a posture combines both preventative controls (predefined and custom organization policies) and detective controls (predefined and custom Security Health Analytics [SHA] modules). Scope this posture to the business unit folder to enforce those controls across the Vertex AI environment.

Why this is correct: You need both preventative guardrails (policies) and detective guardrails (SHA findings) to meet internal security requirements. A posture provides a unified, scalable way to apply and monitor these controls to a specific organizational unit.

Why the other options are weaker:

- A: Assured Workloads focuses on regulatory compliance and data handling, but does not explicitly provide the combined preventative+detective posture for ongoing SCC monitoring. - B: Only adds policies (preventative) without integrating detective SHA modules for ongoing findings. - C: Rego + Workload Manager is not the standard, integrated SCC posture approach and adds complexity without the centralized, unified governance of a SCC posture.

If you want, I can break down the exact components of a SCC posture and how to apply it to a Vertex AI project.

Bengaluru, India

AI Explanation - verified by nanrany on: July 29, 2026
Question 265:
The correct answer is C: Answering billing and account inquiries.
Reasoning:

The AWS Support Concierge is a specialized team within Enterprise Support whose primary role is handling billing and account questions and helping with account setup and getting started.

They are not the source for architecture guidance or development assistance; those services are provided by Technical Account Managers (TAMs) or Solutions Architects as part of Enterprise Support.

For technical support cases, customers typically interact with AWS Support engineers or the TAMs, not the Concierge. The Concierge may route or coordinate related requests, but they don’t provide the technical resolution.

Summary:

What Concierge does: billing and account inquiries, onboarding help, coordination with AWS support.

What Concierge does not do: provide architecture/design guidance or in-depth technical support case resolutions.

Busan, South Korea

AI Explanation - verified by User6293 on: July 20, 2026
Question 1:
Here’s the reasoning for Question 1.

The requirement: the support plan must allow opening new support requests (technical support incidents).

Azure’s Basic plan does not include technical support incidents—only billing/subscription support and access to community support.

Therefore, you must choose a paid plan to open new technical support requests: Developer, Standard, Professional Direct, or Premier.

So, the correct choice is the option that lists the paid support plans (excluding Basic). The typical expectation is: “Developer, Standard, Professional Direct, Premier” as eligible plans. If the option labeled “A” corresponds to that list, then A is correct.
San José, Costa Rica

AI Explanation - verified by User7851 on: August 05, 2026
Question 8:
Correct answer: A — the Microsoft Defender portal
To review the impact of a phishing incident, use the Microsoft Defender portal (security.microsoft.com). It provides security investigation and threat-response tools, including:

Incidents & alerts — see how the phishing activity was detected, correlated, and investigated.

Threat Explorer — search for the malicious messages, recipients, URLs, attachments, and campaign activity.

Email and collaboration reports — assess how many users or messages were affected.

Remediation details — verify whether messages were quarantined, removed, or otherwise contained.

The other portals are not the best choice:

Microsoft 365 admin center — general tenant administration.

Microsoft Entra admin center — identity and access management.

Exchange admin center — mail-flow and Exchange configuration, rather than full phishing-incident investigation.

The key exam distinction is that security incidents and phishing impact analysis belong in Microsoft Defender, not the general administration portals.
Barrie, Canada

AI Explanation - verified by User6346 on: July 16, 2026
Question 22:
Question 22 asks if the given solution meets the goal of reducing admin effort for deploying a large number of VMs weekly (Windows Server 2016 and Ubuntu Linux) and then removing them in the same week.

The proposed solution: use Microsoft Managed Desktop.

Why this does not meet the goal:

- Microsoft Managed Desktop is aimed at managing end-user devices (Windows 10/11) for enterprises, not at provisioning or life-cycle management of Azure VMs. - It does not inherently automate creation/deletion of large VM fleets or support mixed OS images like Windows Server 2016 and Linux.

A better fit would be Azure DevTest Labs (or similar lab services), which is designed to:

- create and manage many VMs for development/testing, - support both Windows and Linux images, - provide templates, auto-shutdown/start, and easier lifecycle management to reduce admin effort.
Answer: No. The MS Managed Desktop approach does not align with the VM lifecycle automation needs; DevTest Labs would be a better choice for this scenario.
Bengaluru, India

AI Explanation - verified by User2839 on: July 28, 2026
Question 6:
Here’s how to think about Question 6.

What it’s asking: In an ARM template, for an Availability Set, what value should you set for the property platformFaultDomainCount to maximize availability during fabric failures or maintenance?

Key concept: Fault domains are groups of hardware that can fail independently. By spreading VMs across more fault domains, you reduce the risk that a single hardware fault takes out many VMs.

Important detail: For managed Availability Sets, the number of fault domains you can use varies by region — typically 2 or 3.

What to choose: Set platformFaultDomainCount to the maximum supported in your region. In most regions this is 3, which gives you the best resilience. If your region only supports 2, you’d use 2.

Answer: 3 (Max Value). In ARM template syntax, this would appear as "platformFaultDomainCount": 3 (within the Availability Set resource).

Mumbai, India

AI Explanation - verified by sivatheanalyst007 on: July 27, 2026
Question 17:
Here’s why Question 17’s answer is E.

Correct answer: only new data added to the eventhouse (option E).

Why: When you enable OneLake availability for the eventhouse, OneLake starts syncing data going forward. It does not automatically backfill or copy over data that already exists in the eventhouse before the enablement.

What that means in practice: Any data added to the eventhouse after you enable OneLake will appear in OneLake (and be visible from OneLake file explorer). Pre-existing data in the eventhouse stays there unless you perform a separate backfill/copy operation outside the standard OneLake enablement.

How to verify: add a small new dataset to the eventhouse after enabling OneLake availability, then check OneLake to confirm the new dataset appears.

Summary: Enabling OneLake availability is incremental for new data; it won’t copy historical data by default.
Singapore, Singapore

AI Explanation - verified by User7395 on: August 05, 2026
Question 8:
Correct answer: D — Parallel implementation.
A parallel implementation runs the old and new systems at the same time. For a month-end process, the organization can continue using the existing system while validating the new system, which helps keep business operations running with minimal downtime.
Why the other options are less suitable:

Cutover: The old system is replaced immediately. It may be fast, but any failure can cause significant disruption.

Phased: The system is introduced in stages, which reduces implementation risk but may prolong the transition.

Pilot: The system is first deployed to a limited group or location. This is useful for testing but does not necessarily minimize downtime for the entire organization.

The trade-off is that parallel implementation requires additional resources because both systems must be operated and their outputs compared. However, when the priority is continuity of a critical process, such as month-end processing, parallel implementation is the best choice.
Jersey City, United States

AI Explanation - verified by User5007 on: July 24, 2026
can you switch to polish?
Tak, moge kontynuowac po polsku.
W czym moge pomóc? Moge:

wyjasnic istniejace pytania z tej strony,

wskazac, dlaczego dana odpowiedz jest poprawna lub bledna (z uwzglednieniem celów egzaminu),

omówic koncepcje zwiazane z Google Cloud AI.

Podaj numer pytania (np. Pytanie 6) i tresc, a ja przeanalizuje je po polsku.
Rybnik, Poland

AI Explanation - verified by User1075 on: July 20, 2026
why not the other answers
Here’s why the other options aren’t the best next step after a new motherboard install with POST failure:

A. Flash the BIOS: Not the first move. A BIOS update can be risky when POST is failing. It’s better to first rule out simpler causes (like RAM seating). BIOS flashing is usually done if the CPU isn’t supported by the current BIOS, which you’d check after RAM seating.

B. Check the CMOS battery: Possible but unlikely the immediate cause. CMOS battery issues would more likely cause BIOS settings to reset or clock problems. RAM seating is a more common, direct fix after a motherboard swap.

D. Reinstall the old motherboard: Not a troubleshooting step for the new system. The goal is to get the new motherboard to POST, not revert to old hardware.

C. Reseat the RAM: Correct because memory seating problems are the most common cause of POST failures after a motherboard swap. Re-seating ensures proper contact in the new board’s slots.

Indpls, United States

AI Explanation - verified by User5630 on: July 29, 2026
Question 51:

The answer: Question 51: C — Request trigger.

Why:

To invoke a Logic App from an external service (here, the Azure Application Insights availability test) you need a trigger that starts on an incoming HTTP request. In Logic Apps, this is the trigger that exposes an HTTP endpoint — typically shown as "When a HTTP request is received" and referred to as the Request trigger.

The other options aren’t appropriate:

- HTTP webhook trigger is for webhook-style callbacks with a handshake and isn’t the standard inbound trigger for a simple HTTP POST. - HTTP trigger is not the commonly exposed inbound trigger in Logic Apps; the inbound trigger is the Request (When a HTTP request is received). - ApiConnection trigger is used to start a flow via a specific connector, not for generic external HTTP events.
How it works in practice:

Create a Logic App with the Request trigger to generate a URL.

Configure Application Insights to post the degraded-availability payload to that URL.

Add actions (e.g., send email, follow-up) in the Logic App to notify the customer services team.

Mumbai, India

AI Explanation - verified by User1935 on: August 02, 2026
Question 7:
Question 7 asks which Azure web tier plan meets these requirements:

Custom domains

10 GB storage per web app

Each app runs on dedicated compute

Load balancing between instances is included

Costs minimized

Answer: A (Standard)

Why:

- Custom domains: All paid tiers support custom domains, but you still need features like dedicated compute and load balancing. - Storage: Free/Shared have very limited resources; 10 GB per app exceeds what those tiers realistically provide. Standard offers ample storage for these apps. - Dedicated compute: Free and Shared run on shared compute. Basic and Standard provide dedicated compute, with Standard being the lowest tier that meets the other requirements. - Load balancing: Built-in load balancing across instances is included in Standard (and higher). Free/Shared don’t meet this production-style load-balancing expectation. - Cost: Basic is cheaper than Standard but lacks one or more required capabilities (notably robust load balancing and sufficient storage). Standard is the lowest tier that satisfies all the criteria.

Note: In App Service, multiple apps in the same plan share compute resources. If you truly need isolation per app, you’d typically use separate App Service Plans, which increases cost beyond a single Standard plan.

Melbourne, Australia

AI Explanation - verified by nanrany on: July 29, 2026
Question 189:
Here’s the explanation for question 189.

Correct answer: D. AWS CloudShell

Why: AWS CloudShell is a browser-based shell that runs in the AWS Console and uses the current authenticated session. It provides a pre-authenticated, ready-to-use command-line environment directly from the Management Console, with AWS CLI and common tools installed.

Why the other options aren’t correct:

- AWS API is an API endpoint, not a browser-based, pre-authenticated shell launched from the console. - AWS Lightsail is a simple VM/service and isn’t designed as a pre-authenticated browser shell launched from the console. - AWS Cloud9 is a browser-based IDE, which requires creating and managing environments and isn’t the same “pre-authenticated shell launched directly from the console.”
Busan, South Korea

AI Explanation - verified by User5640 on: August 10, 2026
You have been tasked with making sure that users who connect to Azure AD via the internet from an unidentified IP address, are automatically encouraged to change passwords. Solution: You configure the use of Azure AD Identity Protection. Does the solution meet the goal?
Yes.
Azure AD Identity Protection can detect risky sign-ins, including sign-ins from unfamiliar or suspicious IP addresses. You can configure a sign-in risk policy to require remediation, such as:

Multi-factor authentication

A secure password change for users whose credentials may be compromised

The policy must be configured with the appropriate risk level and user scope. Identity Protection does not treat every unknown IP as automatically malicious; it evaluates the sign-in risk before applying the policy.
The service is now called Microsoft Entra ID Protection.
Sandton, South Africa

AI Explanation - verified by User3980 on: July 30, 2026
Question 71:
Question 71: Defects associated with project deliverables have been reported. The project team needs to find the cause of the defects. Which of the following tools should the project team use to find the cause? Ishikawa diagram (option C)

Why Ishikawa:

- It’s a root cause analysis tool (fishbone diagram) that helps identify multiple potential causes across categories (e.g., People, Process, Tools, Materials, Environment) and trace them to the underlying problem.

Why not the others:

- Kanban board: visualizes workflow and work in progress, not root cause analysis. - Pareto chart: shows the most frequent issues, useful for prioritization but does not directly identify root causes. - Decision tree: aids decision-making under uncertainty, not primarily for root-cause analysis.
Example: If defects are traced to a coding defect, the Ishikawa diagram might explore categories like methods (coding standards), personnel (training), machines (build tools), materials (libraries), and environment (test data). This helps pinpoint the fundamental cause and guide fixes.
Olive Branch, United States

AI Explanation - verified by User9493 on: July 22, 2026
Question 14:
Question 14 asks how to obtain remote access to a network appliance that has failed to start up properly.

Answer: D) Out-of-band management

Why this is correct:

- When a device isn’t booting, in-band methods like SSH or a web GUI aren’t available because the OS isn’t up. - Out-of-band management uses a dedicated management interface (e.g., BMC/IPMI, iLO, DRAC) on a separate network path. This allows you to access the device’s console, view POST/BIOS output, and perform power cycling or recovery actions even if the main network path is down. - It provides remote console access, remote power control, and configuration recovery capabilities independent of the device’s primary OS.

Why the other options aren’t suitable here:

- Crash cart: physical presence required; not remote. - Jump box: a remote intermediary, but it still relies on the device being reachable and booted; not specifically for a startup failure. - Secure Shell: requires the device to be up and reachable over the network; not usable if it failed to start.
In short, use out-of-band management to recover or administer a device that won’t boot.
Essex, United States

AI Explanation - verified by User6589 on: August 03, 2026
Question 8:
Question 8 asks which improvement opportunity found at baseline can contribute to process maturity and strengthen the customer’s overall security posture.

Post-Incident Review (A) is the after-action step used to capture lessons learned, identify process gaps, and drive improvements across the incident response lifecycle. This directly supports process maturity and a stronger security posture.

Incident Analysis (D) focuses on investigating root causes during/after an incident, which is important but more of a diagnostic activity than a standalone baseline improvement to maturity.

Fast Eradication (B) and Incident Containment (C) are tactical actions aimed at stopping the incident, not on maturing the process.

Note: The page lists D as the answer and the reference link seems unrelated to this question, which suggests an error in the provided key. Based on typical ServiceNow security incident objectives around continuous improvement, Post-Incident Review (A) is the better fit.
Hyderabad, India

AI Explanation - verified by User4710 on: August 12, 2026
Question 2:
Correct answer: D — The ability to support batch and streaming workloads
Delta Lake provides a reliable storage layer built on data files and a transaction log. It supports both:

Batch processing, such as scheduled data transformations

Streaming processing, such as continuously ingesting events

Both workloads can read from and write to the same Delta table, helping unify batch and streaming pipelines.
Why the other options are less appropriate:

Manipulating data with multiple languages: Mainly a benefit of Databricks and Spark.

Real-time notebook collaboration: A workspace/notebook feature.

Alerts for query failures: A Databricks SQL or job-monitoring feature.

Distributing complex operations: Primarily a Spark compute capability.

The key exam association is: Delta Lake supports unified batch and streaming data processing with reliable transactions.
Chennai, India

AI Explanation - verified by User7939 on: August 07, 2026
Question 17:
Correct answer: A, B, C, D, and E.
Azure VM backup can protect:

Windows 10 VMs

Windows Server 2012 or later VMs

Running VMs—they do not need to be shut down

Debian 8.2+ VMs, assuming the specific version is supported

Stopped or deallocated VMs

The key concept is that Azure Backup protects supported Azure VM operating systems whether the VM is running or stopped. Shutting down a VM is not a prerequisite for configuring or performing its backup.
For a running VM, Azure Backup can create a restore point while the VM is online. The consistency level depends on the workload and configuration—for example, application-consistent backups may require additional guest configuration.
Oostburg, The Netherlands

AI Explanation - verified by User9198 on: July 28, 2026
Question 110:
Question 110: The correct answer is D — Amazon S3 File Gateway.
Explanation:

A Storage Gateway File Gateway provides an on-premises NFS/SMB file share that's backed by objects in S3. This is exactly what you need to migrate an NFS workload: applications continue using NFS, while the actual data moves to and is stored in S3.

It can cache frequently accessed data locally for performance, while uploads go to an S3 bucket.

Why the other options aren’t suitable:

- Tape Gateway: for archival, not live file workloads. - Volume Gateway: offers block storage (iSCSI) rather than a file-based interface. - Amazon FSx File Gateway: doesn’t exist; FSx is a separate service (e.g., FSx for Windows File Server) and is not a Storage Gateway file interface.
Busan, South Korea

AI Explanation - verified by phalanndwalivhuwani on: July 20, 2026
Question 40:
Here’s how to interpret Question 40 and why Box 1 and Box 2 are true, while Box 3 is false.

Correct choice: Box 1 = Yes, Box 2 = Yes, Box 3 = No (Answer: A)

Why:

Box 1: Yes – Microsoft Purview Compliance Manager provides a risk-based compliance score that helps you understand your overall compliance posture. The score reflects how well you’re meeting regulatory requirements and internal policies.

Box 2: Yes – It also offers step-by-step remediation guidance through actionable “improvement actions.” These guide you to implement fixes and align controls, often with links to the relevant consoles.

Box 3: No – Compliance Manager is not part of Microsoft Defender. It is part of Microsoft Purview. Defender is a separate security product family, while Compliance Manager is in Purview for compliance risk and remediation.

Key concepts:

Use Purview Compliance Manager to view your compliance score and follow improvement actions.

Improvement actions provide actionable steps and guidance for remediation.

Defender and Purview are distinct services; do not confuse their scopes.

Johannesburg, South Africa

AI Explanation - verified by User6290 on: July 20, 2026
Question 3:
Question 3 asks: In the Create Bulk Cases module, which Filter by options are available in the dropdown? (Choose four.)

According to the provided answer key, the four options are: Upload File, User Criteria, HR Profiles, HR Criteria (options B, D, F, G).

What each means:

Upload File: Filter by a file you upload (e.g., CSV) that provides data for bulk case creation.

User Criteria: Filter by criteria related to users (e.g., role, department) to select target users.

HR Profiles: Filter by HR profile data associated with users.

HR Criteria: Filter by HR-related criteria or attributes.

The other options (Document Template, HR Service Template, HR Template) are more about templates, not typical filtering criteria in this context.
Note: Some communities note that exam answer keys can be incorrect. If you’re unsure, check the official ServiceNow documentation for Create Bulk Cases in your version.
Kolkata, India

AI Explanation - verified by User5630 on: July 29, 2026
Question 16:
Question 16 asks how to make the web app send telemetry at a fixed sampling rate.

Correct answer: B — From the code repository, modify Startup.cs (in ASP.NET Core you configure the Application Insights SDK in code, not with ApplicationInsights.config).

Why the other options aren’t correct:

- Modify ApplicationInsights.config: used in older ASP.NET apps, not typical for ASP.NET Core. - From AppInsights1, modify Usage/cost settings: this doesn’t change sampling behavior. - From AppInsights1, configure Continuous export: export settings affect data export, not sampling.

How to implement fixed-rate sampling (ASP.NET Core):

- In your Startup/Program code, configure the telemetry pipeline to use fixed-rate sampling and disable adaptive sampling. - Example (ASP.NET Core 6+ style after WebApplication.CreateBuilder):
var builder = WebApplication.CreateBuilder(args); builder.Services.Configure<TelemetryConfiguration>(telemetryConfiguration => { var builder = telemetryConfiguration.DefaultTelemetrySink.TelemetryProcessorChainBuilder; double fixedSamplingPercentage = 10; // 10% sampling builder.UseSampling(fixedSamplingPercentage); }); builder.Services.AddApplicationInsightsTelemetry(new ApplicationInsightsServiceOptions { EnableAdaptiveSampling = false }); var app = builder.Build();
- This sets a fixed sampling rate (e.g., 10%) for telemetry sent to AppInsights1.

Key concep
Mumbai, India

AI Explanation - verified by User3779 on: July 24, 2026
Question 121:

The correct answer: C. Tokenization

Why:

- Tokenization replaces sensitive data (like credit card numbers) with a non-value token. The actual data lives in a secure vault, not in the database, so a breach of the database won’t expose real card data. - This approach is widely used for payment card data and reduces exposure risk and PCI scope because the tokens themselves are meaningless outside the tokenization context.

Why the others are weaker:

- Hashing: irreversible, so you can’t retrieve the actual card number to process a transaction. - Obfuscation: typically not secure enough; can often be reversed. - Masking: only hides data in display; the actual data remains stored elsewhere and vulnerable.
Vila Velha, Brazil

AI Explanation - verified by GenericUser9919 on: July 30, 2026
Question 37:
The correct choices are: App Tag and App Score.
Why:

App Tag: a label that reflects the app’s enterprise readiness based on its Cloud Confidence Index (CCI) rating. It lets you quickly group apps (e.g., Poor, Average, Good) for policy decisions.

App Score: a numeric value representing the CCI score. It provides a precise threshold you can use to enforce policies (e.g., block apps with Score below 50).

How this is used in real-time protection:

You can create policies that block or restrict apps with a Poor App Tag or with an App Score under a chosen threshold.

These two aspects let you filter cloud apps by CCI ratings during policy evaluation.

Why the other options aren’t used here:

CCL isn’t described as a real-time policy facet in this scenario.

GDPR Readiness is a separate consideration and not the two CCI aspects used for real-time protection policy in this context.

Charleston, United States

AI Explanation - verified by User1790 on: August 06, 2026
Question 5:
Correct answer: C — Pull
A Databricks Repo can be connected to a central Git repository. When a colleague commits and syncs changes to that central repository, the data engineer uses Pull to retrieve those changes into their local Databricks Repo.
Git operations:

Pull: Downloads changes from the central repository and updates the local repository.

Push: Sends local commits to the central repository.

Commit: Records local changes in the repository’s history.

Merge: Combines changes from different branches.

Clone: Creates a new local copy of a repository.

The key distinction is direction: changes are moving from the central Git repository to the data engineer’s Databricks Repo, so the operation is Pull.
Pune, India

AI Explanation - verified by User9569 on: August 02, 2026
Question 3:

Correct answer: Power BI dataset

Explanation:

The scenario says the data model already contains several measures and you want to create a report from the existing data with minimal development. Reusing a published Power BI dataset lets you inherit the existing measures and data model, so you don’t have to recreate them.

How to do it: In Power BI Desktop, use the connector Power BI datasets (Get Data > Power BI datasets) and connect to the existing dataset, then build visuals using the pre-defined measures.

Why the other options are less suitable:

- SharePoint folder would require importing the Excel data again and rebuilding measures. - Power BI dataflows are for data ingestion pipelines, not for leveraging an existing report’s data model directly. - Excel workbook would reimport data and likely require re-creating the model and measures.
Kuala Lumpur, Malaysia

AI Explanation - verified by User7777 on: July 31, 2026
Question 44:

Answer: B — Improve employee training and awareness

Why this is correct:

- The issue is users unintentionally exposing sensitive data via file sharing. Training and awareness directly address the human factor, teaching users how to identify sensitive data, share appropriately, and follow policy-driven practices. This reduces misconfigurations and careless sharing.

Why the other options are less effective here:

- A) Implement step-up authentication for administrators targets access control for admins, not the regular user behavior causing data leaks. - C) Increase password complexity helps against credential theft but doesn’t prevent missharing of files. - D) Deploying MDM focuses on device management, not the core problem of user-driven data exposure through file sharing.

Quick practical note:

- In practice, you can combine training with policies and technical controls like DLP or file-sharing permissions to further reduce leakage, but training addresses the primary risk in this scenario.
Subang Jaya, Malaysia

AI Explanation - verified by User2505 on: August 10, 2026
Question 19:
Correct answer: A — Provide labeled data with the prompt field and the completion field.
Fine-tuning adapts a foundation model to a company’s specific task, style, or domain. The training examples need to show:

prompt: the input or instruction given to the model

completion: the desired response

For example:

json {"prompt":"Summarize this customer complaint.","completion":"The customer reports a delayed shipment."}

The model learns the relationship between the prompt and the expected completion, allowing it to produce more accurate responses for similar requests.
Why the other options are incorrect:

B: A .txt file containing CSV-style lines is not automatically a valid fine-tuning dataset. The required structure and format must match the selected Bedrock model.

C: Provisioned Throughput provides dedicated inference capacity; it does not fine-tune a model.

D: Journals and textbooks might support domain adaptation or pretraining, but they do not provide the prompt-to-answer examples typically required for supervised fine-tuning.

London, United Kingdom

AI Explanation - verified by bill.dailey on: July 26, 2026
Question 356:
Answer: B — Rootkit
Explanation:

Side loading means installing software from sources outside official or trusted repositories.

This process can bring in malicious payloads because the installed software hasn’t been vetted. A rootkit is a type of malware that hides its presence and maintains privileged access, making it a prime risk when using untrusted sources.

Why not the other options:

- User impersonation is a social engineering tactic, not a direct payload introduced by the install itself. - On-path attack is a network interception (man-in-the-middle), not something installed during side loading. - Buffer overflow is a vulnerability in code; while it could be exploited, “introduced at the time of side loading” implies the install itself bringing in the malicious software, which points to a rootkit.
Mitigation tips (brief):

Only install apps from trusted sources; verify signatures.

Use application whitelisting and security controls to block sideloading.

Keep systems updated and use anti-malware tools capable of detecting rootkits.

Virginia Beach, United States

AI Tutor 👋 I’m here to help!

Cisco 350-901 Exam Actual Questions Developing Applications using Cisco Core Platforms and APIs (DEVCOR) (Page 4 )

QUESTION: 11

QUESTION: 12

Explanation:

QUESTION: 13

QUESTION: 14

QUESTION: 15

Cisco 350-901 Exam Actual Questions
Developing Applications using Cisco Core Platforms and APIs (DEVCOR) (Page 4 )