How should I use these Optimizing performance and cost practice questions?

Read each scenario carefully and choose your answer before revealing the explanation. Then check why your choice was right or wrong. Repeat until the reasoning feels automatic.

Can I practise just Optimizing performance and cost questions in a focused session?

Yes — use the session launcher on this page to start a 10-, 20-, 30- or 50-question session drawn entirely from the Optimizing performance and cost domain.

PCDOE · topic practice

Optimizing performance and cost practice questions

Practise Google Professional Cloud DevOps Engineer Optimizing performance and cost practice questions — original exam-style scenarios with answer choices, explanations, and analysis of common mistakes.

Courseiva uses original exam-style practice questions designed for learning and revision. The goal is to understand the concepts, recognise exam patterns, and improve through explanations — not memorise copied exam dumps.

Reviewed byJohnson Ajibi· MSc IT Security

20 questionsDomain: Optimizing performance and cost

Practice 10 questions Browse domain →

What the exam tests

What to know about Optimizing performance and cost

QoS questions usually test traffic classification (DSCP/CoS), queuing strategies, policing vs shaping and where each mechanism is applied.

DSCP and CoS marking and trust boundaries.

Queuing mechanisms: FIFO, WFQ, CBWFQ, LLQ.

Policing (drops excess) vs shaping (delays excess).

Where QoS is applied: ingress, egress, LAN, WAN.

Watch out for

Common Optimizing performance and cost exam traps

▸Policing drops traffic; shaping buffers it — they solve different problems.
▸DSCP EF (46) is for voice; AF classes are for data tiers.
▸QoS must be configured end-to-end to be effective.
▸CoS values are Layer 2; DSCP is Layer 3.

Practice set

Optimizing performance and cost questions

20 questions · select your answer, then reveal the explanation

Question 1hardmultiple choice

Read the full Optimizing performance and cost explanation →

A company runs a batch processing workload on Compute Engine that runs for 3 hours every night. They want to minimize costs while ensuring the job completes reliably. Which recommendation should they follow?

Trap 1: Use sole-tenant nodes to isolate the workload.

Sole-tenant nodes are for compliance or licensing, not cost savings.

Trap 2: Use standard (on-demand) VMs and enable sustained use discounts.

Sustained use discounts require running for a month; nightly 3-hour jobs don't benefit much.

Trap 3: Purchase 1-year committed use discounts for the VMs.

Committed use discounts require a 1-year commitment and are not cost-effective for short nightly jobs.

Study all Optimizing performance and cost common traps →

A
Use sole-tenant nodes to isolate the workload.
Why wrong: Sole-tenant nodes are for compliance or licensing, not cost savings.
B
Use standard (on-demand) VMs and enable sustained use discounts.
Why wrong: Sustained use discounts require running for a month; nightly 3-hour jobs don't benefit much.
C
Use preemptible VMs and design the job to handle interruptions gracefully.
Preemptible VMs are up to 60% cheaper and suitable for fault-tolerant batch jobs.
D
Purchase 1-year committed use discounts for the VMs.
Why wrong: Committed use discounts require a 1-year commitment and are not cost-effective for short nightly jobs.

Full breakdown with real-world context →

Question 2mediummulti select

Read the full Optimizing performance and cost explanation →

A company wants to control costs by limiting the maximum number of Compute Engine instances a project can create. Which TWO methods can achieve this? (Choose two.)

Trap 1: Use IAM conditions to restrict instance creation.

IAM conditions can restrict instance creation based on attributes such as machine type or region, but they cannot enforce a numeric limit on the total number of instances. This method does not control maximum count.

Trap 2: Enable preemptible VMs.

Enabling preemptible VMs reduces cost but does not limit the maximum number of instances that can be created.

Trap 3: Create budget alerts with a threshold at 80% of budget.

Budget alerts notify when spending reaches a threshold but do not prevent instance creation.

Study all Optimizing performance and cost common traps →

A
Use IAM conditions to restrict instance creation.
Why wrong: IAM conditions can restrict instance creation based on attributes such as machine type or region, but they cannot enforce a numeric limit on the total number of instances. This method does not control maximum count.
B
Enable preemptible VMs.
Why wrong: Enabling preemptible VMs reduces cost but does not limit the maximum number of instances that can be created.
C
Create budget alerts with a threshold at 80% of budget.
Why wrong: Budget alerts notify when spending reaches a threshold but do not prevent instance creation.
D
Set resource quotas for compute instances.
Resource quotas set a hard limit on the number of compute instances that can be created in a project, effectively controlling the maximum count.
E
Apply committed use discounts.
Why wrong: Committed use discounts provide cost savings for consistent usage but do not limit the number of instances.

Full breakdown with real-world context →

Question 3mediummultiple choice

Read the full Optimizing performance and cost explanation →

A company stores historical log data in Cloud Storage. The logs are accessed rarely after 30 days. They want to reduce storage costs while maintaining immediate access for occasional audits over the next year. What should they use?

Trap 1: Nearline class storage

Nearline is cheaper than Standard but still more expensive than Coldline for long-term infrequent access.

Trap 2: Coldline class storage with a lifecycle policy to delete after 1…

Coldline has a 90-day minimum storage charge, and deleting after 1 year may not be optimal without transition from a cheaper class.

Trap 3: Archive class storage

Archive class has retrieval delays and costs, not suitable for occasional immediate access.

Study all Optimizing performance and cost common traps →

A
Standard class storage with a lifecycle policy to transition to Nearline after 30 days and to Coldline after 90 days
This leverages different storage classes based on access patterns, optimizing cost while retaining immediate access after 30 days.
B
Nearline class storage
Why wrong: Nearline is cheaper than Standard but still more expensive than Coldline for long-term infrequent access.
C
Coldline class storage with a lifecycle policy to delete after 1 year
Why wrong: Coldline has a 90-day minimum storage charge, and deleting after 1 year may not be optimal without transition from a cheaper class.
D
Archive class storage
Why wrong: Archive class has retrieval delays and costs, not suitable for occasional immediate access.

Full breakdown with real-world context →

Question 4mediummultiple choice

Read the full Optimizing performance and cost explanation →

A company has a steady-state workload of 100 vCPUs running 24/7. They want to get the maximum discount possible without long-term commitment. What discount should they expect?

Trap 1: Committed use discount of up to 57%

Committed use discounts offer higher discounts but require a 1- or 3-year commitment, which the company does not want.

Trap 2: Sustained use discount of up to 20%

The maximum sustained use discount for full-month usage is 30%, not 20%.

Trap 3: No discount available

Sustained use discounts apply automatically without any commitment.

Study all Optimizing performance and cost common traps →

A
Committed use discount of up to 57%
Why wrong: Committed use discounts offer higher discounts but require a 1- or 3-year commitment, which the company does not want.
B
Sustained use discount of up to 20%
Why wrong: The maximum sustained use discount for full-month usage is 30%, not 20%.
C
Sustained use discount of up to 30%
Sustained use discounts automatically provide up to 30% for running instances the entire month.
D
No discount available
Why wrong: Sustained use discounts apply automatically without any commitment.

Full breakdown with real-world context →

Question 5easymultiple choice

Read the full Optimizing performance and cost explanation →

Refer to the exhibit. A budget alert has fired for project dev-123 indicating that the cost has exceeded the budget of $1000. What should the team do next to investigate the cost overrun?

Exhibit

{
  "budgetName": "projects/my-project/budgets/my-budget",
  "costAmount": 1500.00,
  "budgetAmount": 1000.00,
  "alertThresholdExceeded": 1.0,
  "budgetFilter": {
    "projects": ["projects/dev-123"],
    "creditTypesTreatment": "INCLUDE_ALL_CREDITS"
  }
}

Trap 1: Disable all APIs in project dev-123 immediately.

Disabling APIs could disrupt services; investigation should come first.

Trap 2: Set up a Cloud Function to automatically shut down resources when…

Automated shutdown is a remediation step, not an investigation step.

Trap 3: View the billing account's cost table in the Cloud Console.

The console shows aggregated costs but lacks granular breakdown to pinpoint the cause.

Study all Optimizing performance and cost common traps →

A
Disable all APIs in project dev-123 immediately.
Why wrong: Disabling APIs could disrupt services; investigation should come first.
B
Open a BigQuery query on the billing export table, filtering by project 'dev-123' and service.
BigQuery billing exports provide detailed cost data for root cause analysis.
C
Set up a Cloud Function to automatically shut down resources when the budget is exceeded.
Why wrong: Automated shutdown is a remediation step, not an investigation step.
D
View the billing account's cost table in the Cloud Console.
Why wrong: The console shows aggregated costs but lacks granular breakdown to pinpoint the cause.
E
Create a new budget with a lower threshold to get alerted earlier.
Why wrong: This does not help investigate; it's a reactive measure.

Full breakdown with real-world context →

Question 6mediummultiple choice

Read the full Optimizing performance and cost explanation →

A company is using Google Kubernetes Engine (GKE) with multiple node pools. They notice that their monthly costs are higher than expected. Upon review, they find that several preemptible VMs are being recreated frequently, leading to sustained usage costs. What is the most cost-effective solution to reduce costs?

Trap 1: Purchase committed use discounts for the preemptible VMs.

Committed use discounts are not available for preemptible VMs.

Trap 2: Increase the number of preemptible VMs to spread the workload.

Adding more preemptible VMs would increase costs, not reduce them.

Trap 3: Enable sustained use discounts for the existing VMs.

Sustained use discounts are automatically applied and don't require enabling.

Study all Optimizing performance and cost common traps →

A
Purchase committed use discounts for the preemptible VMs.
Why wrong: Committed use discounts are not available for preemptible VMs.
B
Increase the number of preemptible VMs to spread the workload.
Why wrong: Adding more preemptible VMs would increase costs, not reduce them.
C
Enable sustained use discounts for the existing VMs.
Why wrong: Sustained use discounts are automatically applied and don't require enabling.
D
Migrate to Spot VMs, which have a lower price and no maximum runtime.
Spot VMs are the recommended replacement for preemptible VMs and offer lower costs without the 24-hour limit.

Full breakdown with real-world context →

Question 7easymultiple choice

Read the full Optimizing performance and cost explanation →

A company uses Cloud Storage to store archival data. They want to minimize storage costs while maintaining availability. Which storage class should they use?

Trap 1: Nearline storage class.

Nearline is for data accessed less than once a month, but Archive is cheaper.

Trap 2: Standard storage class.

Standard is for frequently accessed data, not cost-efficient for archival.

Trap 3: Coldline storage class.

Coldline is for data accessed less than once a quarter, but Archive is cheaper.

Study all Optimizing performance and cost common traps →

A
Nearline storage class.
Why wrong: Nearline is for data accessed less than once a month, but Archive is cheaper.
B
Standard storage class.
Why wrong: Standard is for frequently accessed data, not cost-efficient for archival.
C
Archive storage class.
Archive is the lowest-cost storage class for long-term archival data.
D
Coldline storage class.
Why wrong: Coldline is for data accessed less than once a quarter, but Archive is cheaper.

Full breakdown with real-world context →

Question 8mediummultiple choice

Read the full Optimizing performance and cost explanation →

A company is using BigQuery for analytics and wants to control costs. They have many queries that scan large amounts of data. Which approach is most effective in reducing query costs?

Trap 1: Switch to flat-rate pricing to cap costs.

Flat-rate pricing caps total cost but does not reduce per-query cost; it may be more expensive for low usage.

Trap 2: Reserve BigQuery slots for dedicated capacity.

Reserving slots is for dedicated capacity, not cost reduction; it may increase costs.

Trap 3: Use clustering to organize data within partitions.

Clustering improves performance but does not directly reduce the amount of data scanned.

Study all Optimizing performance and cost common traps →

A
Switch to flat-rate pricing to cap costs.
Why wrong: Flat-rate pricing caps total cost but does not reduce per-query cost; it may be more expensive for low usage.
B
Partition tables by date and use partition pruning in queries.
Partitioning limits the data scanned, reducing query costs.
C
Reserve BigQuery slots for dedicated capacity.
Why wrong: Reserving slots is for dedicated capacity, not cost reduction; it may increase costs.
D
Use clustering to organize data within partitions.
Why wrong: Clustering improves performance but does not directly reduce the amount of data scanned.

Full breakdown with real-world context →

Question 9hardmultiple choice

Read the full Optimizing performance and cost explanation →

A company uses Cloud CDN to deliver content globally. They notice increasing egress costs. Which change will most effectively reduce egress costs?

Trap 1: Switch to a premium tier network for lower egress rates.

Premium tier has higher per-GB egress costs than standard tier.

Trap 2: Enable gzip compression for all responses.

Compression reduces data size but Cloud CDN charges per byte after compression; still, egress costs are based on the compressed size, so savings are limited.

Trap 3: Use Cloud Armor to block malicious traffic.

Cloud Armor protects against attacks but does not reduce legitimate egress costs.

Study all Optimizing performance and cost common traps →

A
Switch to a premium tier network for lower egress rates.
Why wrong: Premium tier has higher per-GB egress costs than standard tier.
B
Enable gzip compression for all responses.
Why wrong: Compression reduces data size but Cloud CDN charges per byte after compression; still, egress costs are based on the compressed size, so savings are limited.
C
Use Cloud Armor to block malicious traffic.
Why wrong: Cloud Armor protects against attacks but does not reduce legitimate egress costs.
D
Configure Cloud CDN to cache more content and increase cache hit ratio.
Higher cache hit ratio reduces the amount of data fetched from the origin, lowering egress costs.

Full breakdown with real-world context →

Question 10mediummulti select

Read the full Optimizing performance and cost explanation →

A company is using Google Cloud and wants to monitor and control costs. Which TWO actions should they take? (Choose two.)

Trap 1: Disable all unnecessary APIs at the organization level.

Disabling APIs globally may affect other projects; better to manage per project.

Trap 2: Export billing data to BigQuery for detailed analysis.

Exporting to BigQuery is for analysis, not direct cost control.

Trap 3: Grant billing account access to all project owners.

Granting billing access to all owners increases risk of accidental spending.

Study all Optimizing performance and cost common traps →

A
Set up budget alerts to notify when spending exceeds thresholds.
Budget alerts help monitor and control costs by providing notifications.
B
Use labels to categorize resources and track costs by team.
Labels enable cost allocation and reporting, helping control costs.
C
Disable all unnecessary APIs at the organization level.
Why wrong: Disabling APIs globally may affect other projects; better to manage per project.
D
Export billing data to BigQuery for detailed analysis.
Why wrong: Exporting to BigQuery is for analysis, not direct cost control.
E
Grant billing account access to all project owners.
Why wrong: Granting billing access to all owners increases risk of accidental spending.

Full breakdown with real-world context →

Question 11hardmulti select

Read the full Optimizing performance and cost explanation →

A company runs a web application on Google Kubernetes Engine (GKE) with multiple services. They want to reduce costs without impacting performance. Which THREE actions should they take? (Choose three.)

Trap 1: Enable cluster autoscaling and manually scale nodes based on peak…

Manual scaling is not efficient; node auto-provisioning is better for cost.

Trap 2: Deploy a service mesh like Istio to optimize traffic routing.

Service mesh adds overhead and cost, not reducing costs.

Study all Optimizing performance and cost common traps →

A
Enable cluster autoscaling and manually scale nodes based on peak load.
Why wrong: Manual scaling is not efficient; node auto-provisioning is better for cost.
B
Deploy a service mesh like Istio to optimize traffic routing.
Why wrong: Service mesh adds overhead and cost, not reducing costs.
C
Enable node auto-provisioning to automatically adjust node pools.
Node auto-provisioning ensures the cluster uses the right size and type of nodes.
D
Right-size CPU and memory requests and limits for each service.
Right-sizing avoids over-provisioning and reduces node resource waste.
E
Use preemptible VMs for stateless, fault-tolerant workloads.
Preemptible VMs are cheaper and suitable for fault-tolerant stateless services.

Full breakdown with real-world context →

Question 12hardmultiple choice

Read the full Optimizing performance and cost explanation →

You are a DevOps engineer at a large e-commerce company that runs its production workloads on Google Kubernetes Engine (GKE) in the us-central1 region. The cluster has 500 nodes, each with 8 vCPUs and 32 GB of memory, and uses preemptible VMs for cost savings. Over the past month, the monthly GKE cost has increased by 30% unexpectedly. Upon reviewing the billing reports, you notice a significant spike in Compute Engine costs, specifically for 'Sustained Use Discount' line items, but the total cost is higher than expected. You also observe that the cluster's node utilization is inconsistent, with some nodes running at 90% CPU and memory while others are below 20%. Your team has been deploying stateless microservices and using Cluster Autoscaler with default settings. The application traffic is variable but predictable, with peaks on weekends. You need to reduce the GKE costs without impacting performance. What should you do?

Trap 1: Increase the minimum number of nodes in the cluster to 300 and use…

Increasing minimum nodes and using larger machines increases costs and may lead to underutilization.

Trap 2: Switch the cluster to a single-zone configuration and reduce the…

Single-zone reduces availability and may cause downtime, and reducing nodes may not handle peak loads.

Trap 3: Reduce the number of preemptible VMs to 30% and use only on-demand…

Reducing preemptible VMs increases cost because on-demand VMs are more expensive.

Study all Optimizing performance and cost common traps →

A
Enable node auto-provisioning and migrate baseline workloads to nodes covered by committed use discounts (1-year or 3-year).
Node auto-provisioning optimizes resource allocation based on pod requirements, and CUDs provide significant savings for stable workloads.
B
Increase the minimum number of nodes in the cluster to 300 and use a larger machine type to reduce the number of pods per node.
Why wrong: Increasing minimum nodes and using larger machines increases costs and may lead to underutilization.
C
Switch the cluster to a single-zone configuration and reduce the number of nodes to 200 to lower base costs.
Why wrong: Single-zone reduces availability and may cause downtime, and reducing nodes may not handle peak loads.
D
Reduce the number of preemptible VMs to 30% and use only on-demand VMs for the remaining nodes to improve reliability.
Why wrong: Reducing preemptible VMs increases cost because on-demand VMs are more expensive.

Full breakdown with real-world context →

Question 13hardmultiple choice

Read the full Optimizing performance and cost explanation →

A company runs a critical application on Google Kubernetes Engine (GKE) with 3 nodes (e2-standard-4). To reduce costs, the team is considering right-sizing the nodes. The application is latency-sensitive and experiences periodic traffic spikes. What is the most cost-effective approach that maintains performance during spikes?

Trap 1: Use a single larger node (n2-standard-8) to reduce node count and…

This creates a single point of failure and may not handle spikes efficiently.

Trap 2: Switch to 3 n2-standard-2 nodes to reduce vCPU and memory, and rely…

This may cause resource contention during spikes as the smaller nodes have less capacity.

Trap 3: Keep current node size but use committed use discounts for 1 year…

Committed use discounts require upfront commitment and may not be optimal if right-sizing is possible.

Study all Optimizing performance and cost common traps →

A
Use a single larger node (n2-standard-8) to reduce node count and network overhead.
Why wrong: This creates a single point of failure and may not handle spikes efficiently.
B
Switch to 3 n2-standard-2 nodes to reduce vCPU and memory, and rely on horizontal pod autoscaling.
Why wrong: This may cause resource contention during spikes as the smaller nodes have less capacity.
C
Create a node pool with smaller nodes (e2-standard-2) using preemptible VMs and enable cluster autoscaler.
Preemptible VMs are cheaper, and autoscaler adds nodes during spikes, maintaining performance.
D
Keep current node size but use committed use discounts for 1 year to reduce per-hour cost.
Why wrong: Committed use discounts require upfront commitment and may not be optimal if right-sizing is possible.

Full breakdown with real-world context →

Question 14mediummulti select

Read the full Optimizing performance and cost explanation →

A DevOps team is analyzing Google Cloud costs and notices that spending on BigQuery has increased significantly. They want to reduce costs without impacting ongoing analytical workloads. Which TWO actions should they take? (Choose two.)

Trap 1: Switch to on-demand pricing to pay only for queries run.

On-demand may be more expensive for high-volume workloads; flat-rate is better.

Trap 2: Enable column-level security to restrict access to sensitive data.

This is a security measure, not directly cost-saving.

Trap 3: Delete unused datasets to reduce storage costs.

Deleting datasets removes data and may impact analytics; better to archive old data.

Study all Optimizing performance and cost common traps →

A
Switch to on-demand pricing to pay only for queries run.
Why wrong: On-demand may be more expensive for high-volume workloads; flat-rate is better.
B
Enable column-level security to restrict access to sensitive data.
Why wrong: This is a security measure, not directly cost-saving.
C
Set custom cost controls like query quotas and maximum bytes billed per query.
Limits prevent expensive queries from running unbounded.
D
Delete unused datasets to reduce storage costs.
Why wrong: Deleting datasets removes data and may impact analytics; better to archive old data.
E
Implement flat-rate pricing with reservations for consistent workloads.
Flat-rate pricing caps costs for predictable usage, avoiding per-query charges.

Full breakdown with real-world context →

Question 15easymultiple choice

Read the full Optimizing performance and cost explanation →

Refer to the exhibit. The output shows a recommendation from the Cloud Cost Optimization recommender for an instance in us-central1-a. The instance is a production web server that consistently runs at 25% CPU utilization during peak hours. What should the DevOps engineer do to implement this recommendation with minimal risk?

Network Topology

Trap 1: Ignore the recommendation because the instance is production and…

The recommendation is based on low utilization, and a machine type change can be done with a brief restart.

Trap 2: Add a second n2-standard-4 instance behind a load balancer to…

This would increase cost, not reduce it.

Trap 3: Use a custom machine type with 4 vCPUs and 32 GB memory to ensure…

The recommendation is for n2-standard-4 (16 GB RAM), and over-provisioning RAM may not be needed.

Study all Optimizing performance and cost common traps →

A
Stop the instance, change the machine type to n2-standard-4, and start it again.
This directly implements the recommendation with minimal risk.
B
Ignore the recommendation because the instance is production and any change might cause downtime.
Why wrong: The recommendation is based on low utilization, and a machine type change can be done with a brief restart.
C
Add a second n2-standard-4 instance behind a load balancer to distribute load.
Why wrong: This would increase cost, not reduce it.
D
Use a custom machine type with 4 vCPUs and 32 GB memory to ensure enough RAM.
Why wrong: The recommendation is for n2-standard-4 (16 GB RAM), and over-provisioning RAM may not be needed.

Full breakdown with real-world context →

Question 16mediumdrag order

Read the full Optimizing performance and cost explanation →

Arrange the steps to troubleshoot a high latency issue on a Google Cloud HTTP(S) Load Balancer.

Drag steps to the numbered slots on the right, or tap a step then tap a slot.

Steps

Order

1Step 1

2Step 2

3Step 3

4Step 4

5Step 5

Trap 1: Analyze metrics, then review logs, then check health checks, then…

This is incorrect because health checks should be performed first to rule out backend health issues before diving into metrics and logs.

Trap 2: Check health checks, then identify the problematic backend from…

This is incorrect because taking action before analyzing metrics may not address the root cause; metrics are needed to quantify and confirm the issue after identification.

Trap 3: Review logs, then analyze metrics, then identify the problematic…

This is incorrect because health checks are a crucial first step; skipping them may miss backend health issues that could cause latency.

Study all Optimizing performance and cost common traps →

A
Check health checks, then review load balancer logs, then analyze metrics, then identify the problematic backend, then take corrective action.
This is the correct order because you first ensure backends are healthy via health checks, then examine logs for details, use metrics to quantify latency, identify the specific culprit, and finally take action to resolve the issue.
B
Analyze metrics, then review logs, then check health checks, then identify the problematic backend, then take corrective action.
Why wrong: This is incorrect because health checks should be performed first to rule out backend health issues before diving into metrics and logs.
C
Check health checks, then identify the problematic backend from logs, then take corrective action, then analyze metrics, then verify.
Why wrong: This is incorrect because taking action before analyzing metrics may not address the root cause; metrics are needed to quantify and confirm the issue after identification.
D
Review logs, then analyze metrics, then identify the problematic backend, then take corrective action.
Why wrong: This is incorrect because health checks are a crucial first step; skipping them may miss backend health issues that could cause latency.

Full breakdown with real-world context →

Question 17mediumdrag order

Read the full Optimizing performance and cost explanation →

Order the steps to respond to a Google Cloud security incident involving a compromised service account key.

Drag steps to the numbered slots on the right, or tap a step then tap a slot.

Steps

Order

1Step 1

2Step 2

3Step 3

4Step 4

5Step 5

Trap 1: Audit logs, then revoke key, then rotate key, then assess impact,…

This is incorrect because auditing before revocation delays containment, allowing the attacker continued access while logs are reviewed.

Trap 2: Rotate key, then revoke key, then audit logs, then assess impact,…

This is incorrect because rotating first without revoking leaves the compromised key active, defeating the purpose of rotation.

Trap 3: Revoke key, then audit logs, then rotate key, then assess impact,…

This is incorrect because delaying rotation after revocation means legitimate users cannot access resources until logs are reviewed, which is unnecessary.

Study all Optimizing performance and cost common traps →

A
Revoke key, then rotate key, then audit logs, then assess impact, then update policies.
This order is correct because immediate revocation stops the threat, rotation restores legitimate access quickly, logging provides forensic data, impact assessment evaluates damage, and policy updates prevent recurrence.
B
Audit logs, then revoke key, then rotate key, then assess impact, then update policies.
Why wrong: This is incorrect because auditing before revocation delays containment, allowing the attacker continued access while logs are reviewed.
C
Rotate key, then revoke key, then audit logs, then assess impact, then update policies.
Why wrong: This is incorrect because rotating first without revoking leaves the compromised key active, defeating the purpose of rotation.
D
Revoke key, then audit logs, then rotate key, then assess impact, then update policies.
Why wrong: This is incorrect because delaying rotation after revocation means legitimate users cannot access resources until logs are reviewed, which is unnecessary.

Full breakdown with real-world context →

Question 18mediummatching

Read the full Optimizing performance and cost explanation →

Match each SRE term to its description.

Drag a concept onto its matching description — or click a concept then click the description.

Concepts

Matches

Actual measured service performance

Target reliability level for a service

Contractual commitment to customers

Allowed downtime before SLO breach

Manual, repetitive operational work

Trap 1: SLI: The acceptable amount of unreliability remaining before the…

Incorrect — this definition actually describes the Error Budget, not the SLI.

Trap 2: SLO: A contract with users that includes consequences.

Incorrect — this definition describes an SLA, not an SLO. SLOs are internal targets, not contracts.

Study all Optimizing performance and cost common traps →

A
SLI: A carefully defined quantitative measure of some aspect of the level of service.
SLI stands for Service Level Indicator and is a specific metric used to measure the service's performance.
B
SLO: A target value or range of values for a service level that is measured by an SLI.
SLO (Service Level Objective) defines the desired state of the service as measured by SLIs.
C
SLA: An explicit or implicit contract with users that includes consequences of meeting (or missing) SLOs.
SLA (Service Level Agreement) is a formal agreement that often includes penalties or rewards based on SLO achievement.
D
Error Budget: The acceptable amount of unreliability remaining before the SLO is violated.
Error Budget is the inverse of the SLO: it's the allowed downtime or failures before breaching the objective.
E
SLI: The acceptable amount of unreliability remaining before the SLO is violated.
Why wrong: Incorrect — this definition actually describes the Error Budget, not the SLI.
F
SLO: A contract with users that includes consequences.
Why wrong: Incorrect — this definition describes an SLA, not an SLO. SLOs are internal targets, not contracts.

Full breakdown with real-world context →

Question 19mediummatching

Read the full Optimizing performance and cost explanation →

Match each cost optimization practice to its description.

Drag a concept onto its matching description — or click a concept then click the description.

Concepts

Matches

Discount for 1- or 3-year resource commitment

Automatic discounts for running instances most of month

Short-lived, low-cost instances for batch jobs

Adjusting machine type to match workload needs

Notifications when spending exceeds thresholds

Trap 1: Committed Use Discounts: Short-lived, lower-cost instances that can…

Incorrect — this describes Preemptible VMs, not Committed Use Discounts.

Trap 2: Preemptible VMs: Discounts for committing to use a certain amount…

Incorrect — this describes Committed Use Discounts, not Preemptible VMs.

Study all Optimizing performance and cost common traps →

A
Committed Use Discounts: Discounts for committing to use a certain amount of resources for 1 or 3 years.
Committed Use Discounts (CUDs) reward users for committing to a specific resource usage level for a 1- or 3-year term, providing significant cost savings.
B
Preemptible VMs: Short-lived, lower-cost instances that can be terminated at any time.
Preemptible VMs are spot instances that are cheaper but can be preempted (terminated) by Google Cloud when resources are needed elsewhere.
C
Sustained Use Discounts: Automatic discounts for running instances for a significant portion of a month.
Sustained Use Discounts (SUDs) are automatically applied when an instance runs for more than 25% of a month, offering incremental savings.
D
Custom Machine Types: Creating VMs with specific vCPU and memory configurations to match workload needs.
Custom machine types allow tailoring vCPU and RAM to avoid overprovisioning and reduce costs.
E
Committed Use Discounts: Short-lived, lower-cost instances that can be terminated at any time.
Why wrong: Incorrect — this describes Preemptible VMs, not Committed Use Discounts.
F
Preemptible VMs: Discounts for committing to use a certain amount of resources for 1 or 3 years.
Why wrong: Incorrect — this describes Committed Use Discounts, not Preemptible VMs.

Full breakdown with real-world context →

Question 20mediummultiple choice

Read the full Optimizing performance and cost explanation →

A company uses Compute Engine with committed use discounts for 1-year. They need to reduce costs further. What should they do?

Trap 1: Use sustained use discounts instead.

Sustained use discounts apply automatically and cannot be used as a replacement for committed use discounts.

Trap 2: Use preemptible VMs for all workloads.

Preemptible VMs can be terminated at any time and are not suitable for all workloads.

Trap 3: Increase committed use discount term to 3 years.

Increasing commitment may provide higher discounts but requires longer commitment and may not address overprovisioning.

Study all Optimizing performance and cost common traps →

A
Use sustained use discounts instead.
Why wrong: Sustained use discounts apply automatically and cannot be used as a replacement for committed use discounts.
B
Use preemptible VMs for all workloads.
Why wrong: Preemptible VMs can be terminated at any time and are not suitable for all workloads.
C
Rightsize their VMs based on recommender recommendations.
Rightsizing reduces resource usage, directly lowering costs.
D
Increase committed use discount term to 3 years.
Why wrong: Increasing commitment may provide higher discounts but requires longer commitment and may not address overprovisioning.

Full breakdown with real-world context →

Continue with 20-question session →

Free account

Track your progress over time

Create a free account to save your results and see which topics improve across sessions.

Focused Optimizing performance and cost sessions

Start a Optimizing performance and cost only practice session

Every question in these sessions is drawn from the Optimizing performance and cost domain — nothing else.

10 questions 20 questions 30 questions 50 questions

Browse all Optimizing performance and cost questions →Mixed PCDOE session

Frequently asked questions

What does the PCDOE exam test about Optimizing performance and cost?: QoS questions usually test traffic classification (DSCP/CoS), queuing strategies, policing vs shaping and where each mechanism is applied.
How should I use these practice questions?: Select your answer before revealing the explanation. Then read why each option is right or wrong — this active recall approach builds retention far faster than re-reading notes.
Can I practise just Optimizing performance and cost questions in a focused session?: Yes — the session launcher on this page draws every question from the Optimizing performance and cost domain. Use a 10-question session first to gauge your baseline, then move to 20 or 30 once the weak spots are clear.
Where can I practise other PCDOE topics?: Use the topic links above to move to related areas, or go back to the PCDOE question bank to see all topics.
Are these real exam questions or dumps?: These are original practice questions written to test the same concepts the PCDOE exam covers. They are not copied from any real exam or dump site.

Optimizing performance and cost only

10 questions 20 questions 30 questions 50 questions

Mixed PCDOE session

Track your progress

A free account saves results across sessions and highlights which topics need work.

Study resources

All PCDOE questions Optimizing performance and cost domain overview PCDOE exam guide

Exam traps to avoid

▸Policing drops traffic; shaping buffers it — they solve different problems.
▸DSCP EF (46) is for voice; AF classes are for data tiers.
▸QoS must be configured end-to-end to be effective.
▸CoS values are Layer 2; DSCP is Layer 3.

Optimizing performance and cost practice questions

What to know about Optimizing performance and cost

Common Optimizing performance and cost exam traps

Optimizing performance and cost questions

A company runs a batch processing workload on Compute Engine that runs for 3 hours every night. They want to minimize costs while ensuring the job completes reliably. Which recommendation should they follow?

A company wants to control costs by limiting the maximum number of Compute Engine instances a project can create. Which TWO methods can achieve this? (Choose two.)

A company stores historical log data in Cloud Storage. The logs are accessed rarely after 30 days. They want to reduce storage costs while maintaining immediate access for occasional audits over the next year. What should they use?

A company has a steady-state workload of 100 vCPUs running 24/7. They want to get the maximum discount possible without long-term commitment. What discount should they expect?

Refer to the exhibit. A budget alert has fired for project dev-123 indicating that the cost has exceeded the budget of $1000. What should the team do next to investigate the cost overrun?

Exhibit

A company uses Cloud Storage to store archival data. They want to minimize storage costs while maintaining availability. Which storage class should they use?

A company is using BigQuery for analytics and wants to control costs. They have many queries that scan large amounts of data. Which approach is most effective in reducing query costs?

A company uses Cloud CDN to deliver content globally. They notice increasing egress costs. Which change will most effectively reduce egress costs?

A company is using Google Cloud and wants to monitor and control costs. Which TWO actions should they take? (Choose two.)

A company runs a web application on Google Kubernetes Engine (GKE) with multiple services. They want to reduce costs without impacting performance. Which THREE actions should they take? (Choose three.)

A DevOps team is analyzing Google Cloud costs and notices that spending on BigQuery has increased significantly. They want to reduce costs without impacting ongoing analytical workloads. Which TWO actions should they take? (Choose two.)

Arrange the steps to troubleshoot a high latency issue on a Google Cloud HTTP(S) Load Balancer.

Order the steps to respond to a Google Cloud security incident involving a compromised service account key.

Match each SRE term to its description.

Match each cost optimization practice to its description.

A company uses Compute Engine with committed use discounts for 1-year. They need to reduce costs further. What should they do?

Track your progress over time

Start a Optimizing performance and cost only practice session

Related PCDOE topic practice pages

Bootstrapping a Google Cloud organization for DevOps practice questions

Applying site reliability engineering practices practice questions