PCDE Practice Exam 63 — 20 Questions

Question 1

You are running a Cloud Run service that performs background image processing tasks after responding to HTTP requests. Users report that tasks are sometimes not completed. You have set CPU to always-on, but you notice the service's concurrency is set to the default of 80. What is the most likely cause of incomplete tasks?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

The concurrency setting of 80 allows too many requests to share the instance's CPU, starving background tasks.. Cloud Run allows up to 80 concurrent requests per instance by default. If the concurrency is too high, the instance's CPU may be overwhelmed, causing background tasks to be preempted or not completed. Setting CPU to always-on ensures CPU is available but does not limit concurrency. Reducing concurrency to a lower value (e.g., 1) ensures each request gets dedicated CPU for background work.

Answer

The service should use Gen2 execution environment for better background processing.

Answer

The CPU always-on setting is not sufficient; you need to enable background CPU.

Answer

You need to increase the request timeout beyond 60 minutes.

Question 2

Refer to the exhibit. You are reviewing a Firestore security rules file. What is the main security flaw in the database schema design that these rules expose?

Accepted Answer

Users can set the visibility field, allowing them to make messages public. The Firestore security rules allow any authenticated user to set the `visibility` field on a message document. This means a user could change the visibility to 'public', making private messages accessible to all users regardless of the intended audience. The rules do not validate that the user setting the visibility is the sender or an admin, exposing a data access control flaw.

Answer

The rules do not protect against brute force attacks

Answer

The senderId field is not indexed

Answer

The delete rule allows admin to delete any message

Question 3

A team wants to reduce toil in their operations. Which two of the following are characteristics of toil according to Google SRE principles? (Choose 2)

Accepted Answer

Work that is manual and repetitive. Toil is manual, repetitive, automatable, devoid of enduring value, and scales linearly with service growth. Work that is creative or strategic is not toil. Tasks that require deep analysis are not toil.

Answer

Work that provides enduring value for the service

Answer

Work that is completely automated already

Answer

Work that requires creative problem-solving

Question 4

Which THREE actions can reduce read latency for a globally distributed Cloud Spanner database?

Accepted Answer

Use read-only replicas (follower reads). Read-only replicas (follower reads) allow Cloud Spanner to serve read requests from non-leader replicas, reducing the distance data must travel and thus lowering read latency for globally distributed users. This is particularly effective when strong consistency is not required, as follower reads can return data that is up to a few seconds stale but much faster to access from a nearby replica.

Answer

Use leader reads

Answer

Use interleaved tables

Question 5

A data analyst is running a BigQuery query that joins multiple tables to generate a BI report. The query is slow and uses many LEFT JOINs. What is the best approach to improve performance without changing the business logic?

Accepted Answer

Denormalize the data using nested repeated fields to avoid joins. Denormalizing data using nested repeated fields in BigQuery reduces the number of JOIN operations, which are expensive in a distributed, columnar storage system. By storing related data in a single table with REPEATED fields, the query avoids shuffling large datasets across slots, directly improving performance while preserving the original business logic.

Answer

Add indexes on the join columns

Answer

Replace LEFT JOINs with INNER JOINs where possible

Answer

Increase the number of BigQuery slots

Question 6

An e-commerce platform uses Cloud SQL for PostgreSQL. They need to run complex reporting queries that join several tables. These queries are slowing down the transactional workload. What should they do?

Accepted Answer

Use read replicas to offload reporting queries.. Read replicas in Cloud SQL for PostgreSQL allow you to offload read-only queries, such as complex reporting joins, from the primary instance. This separation reduces contention for CPU, memory, and I/O resources, preserving transactional performance. Read replicas are asynchronous and can handle heavy analytical workloads without impacting the primary's write path.

Answer

Create materialized views for common reports.

Answer

Change all joins to use subqueries.

Answer

Increase the number of vCPUs on the primary instance.

Question 7

A company runs a production Cloud SQL for PostgreSQL instance used by a web application. The instance experiences intermittent latency spikes during peak hours. You need to diagnose the cause without downtime. Which tool should you use first?

Accepted Answer

Use Cloud SQL Insights to analyze query performance and identify slow queries.. Cloud SQL Insights provides built-in query performance monitoring and diagnostics without requiring any downtime. It surfaces slow queries, lock contention, and resource bottlenecks directly from the PostgreSQL engine, making it the ideal first step to identify the root cause of intermittent latency spikes during peak hours.

Answer

Use Database Migration Service to failover to a read replica.

Answer

Use gcloud sql instances describe to check instance configuration.

Answer

Use VPC Flow Logs to analyze network traffic.

Question 8

A company is deploying a microservice to Cloud Run. They need to ensure that the service can handle sudden traffic spikes without cold starts. The service has a memory requirement of 2 GB and CPU usage is bursty. What should they configure?

Accepted Answer

Set --min-instances to 2 and --cpu-throttling to false.. Setting --min-instances to 2 ensures that at least two container instances are always warm, eliminating cold starts during traffic spikes. Setting --cpu-throttling to false disables CPU throttling, allowing the service to use burst CPU capacity immediately without waiting for allocation, which is ideal for bursty CPU workloads.

Answer

Set --ingress to all and --vpc-connector to a Serverless VPC Access connector.

Answer

Set --max-instances to 100 and --concurrency to 80.

Answer

Set --memory to 2Gi and --cpu to 2.

Question 9

A company uses Cloud Spanner for a global inventory application. They need to add a new index to a large table without downtime. What is the recommended approach?

Accepted Answer

Use CREATE INDEX; Spanner handles it as a non-blocking operation. Cloud Spanner's CREATE INDEX operation is inherently non-blocking; it performs the index backfill as a background process without locking the table or blocking reads and writes. This allows the index to be added to a large table with zero downtime, as Spanner manages the consistency and availability of the data throughout the operation.

Answer

Create the index using ALTER TABLE with 'ONLINE' clause

Answer

Use gcloud spanner databases ddl update with --async flag

Answer

Take a backup, create index on a new instance, and switch traffic

Question 10

Which TWO are best practices for designing a star schema in BigQuery for BI? (Choose two.)

Accepted Answer

Store dimension attributes in a single denormalized dimension table instead of multiple normalized tables.. In BigQuery, storing dimension attributes in a single denormalized dimension table (star schema) reduces the number of joins required in BI queries, improving query performance and simplifying SQL. BigQuery's columnar storage and distributed architecture handle denormalized dimensions efficiently, avoiding the overhead of multiple normalized tables that would require complex joins and slow down analytical queries.

Answer

Partition fact tables by low-cardinality columns like gender.

Answer

Pre-aggregate all measures at every possible grain in the fact table.

Answer

Avoid using joins entirely by storing all data in one wide table.

Question 11

A company has a parent-child relationship in Cloud Spanner. They want to minimize cross-table join latency. What should they use?

Accepted Answer

Interleaved tables

Answer

Cloud Functions

Answer

Stored procedures

Answer

Secondary indexes

Question 12

An application uses Cloud SQL (MySQL) and experiences an increasing number of 'too many connections' errors. The current instance has 4 vCPUs and 15GB of RAM. The application's connection pool is configured for 500 connections. What should the engineer do to resolve this error?

Accepted Answer

Increase the max_connections flag in the Cloud SQL database flags settings. The max_connections in MySQL is typically set based on memory: max_connections = RAM_MB/16. With 15GB RAM = 15360 MB, max_connections = 15360/16 = 960. The application pool of 500 should be fine, but if the error occurs, it might be due to other factors. However, the most common fix is to increase the max_connections parameter via a flag. The correct step is to increase the 'max_connections' flag in Cloud SQL. Reducing connections or using a connection pool won't help if the limit is reached. Increasing vCPUs does not directly change max_connections.

Answer

Configure a Cloud SQL connection pool with a maximum of 100 connections

Answer

Create a read replica to distribute read connections

Answer

Upgrade the instance to have 8 vCPUs to increase connection capacity

Question 13

Which THREE considerations are important when designing a schema for Cloud Firestore to ensure scalability?

Accepted Answer

Design collections to avoid high read/write rates on a single document.. Cloud Firestore scales by distributing data across multiple documents. A single document with high read/write rates creates a hotspot, leading to contention and degraded performance. Designing collections to spread load evenly avoids this bottleneck and ensures linear scalability.

Answer

Nest subcollections up to 10 levels deep to model complex hierarchies.

Answer

Use collection group indexes for all queries to avoid manual index creation.

Question 14

A Cloud Spanner application experiences high write latency on a table with a monotonically increasing primary key. Which schema change will most effectively reduce latency?

Accepted Answer

Modify the primary key to include a hash of the original key as a leading column. Monotonically increasing primary keys in Cloud Spanner cause writes to be concentrated on a single split (hotspotting), leading to high write latency. By modifying the primary key to include a hash of the original key as a leading column, writes are distributed uniformly across all nodes, eliminating the hotspot. Option A (interleaved tables) is a table organization pattern that does not address hotspotting. Option B (secondary index on existing key) does not change the write pattern; writes still go to the same primary key range. Option D (increasing nodes) can improve overall throughput but does not fix the hotspotting; writes remain concentrated on one split, so latency may not improve significantly.

Answer

Convert the table to an interleaved table

Answer

Add a secondary index on the existing key

Answer

Increase the number of nodes in the instance

Question 15

A data scientist runs a complex SQL query on a large BigQuery dataset and receives the above error. The query joins 10 tables and uses multiple window functions. Which action is most likely to resolve the issue?

Accepted Answer

Simplify the query by reducing the number of joins or using a temporary table.. The error is likely a resource or memory exhaustion error caused by the complexity of the query (10 joins and multiple window functions). Simplifying the query by reducing joins or using temporary tables reduces the amount of data shuffled and processed in a single stage, directly addressing the root cause. In BigQuery, complex queries with many joins and window functions can exceed slot memory limits, and breaking them into simpler steps avoids this.

Answer

Apply for a quota increase for concurrent queries.

Answer

Increase the number of slots allocated to the project.

Answer

Use the '--maximum_billing_tier' flag to increase the billing tier.

Question 16

A team is managing a Memorystore for Redis instance that needs to scale to handle increased traffic. They want to ensure high availability and the ability to distribute data across multiple nodes. Which three actions should they take? (Choose THREE.)

Accepted Answer

Enable Redis Cluster on the instance to shard data across multiple nodes. Memorystore for Redis offers vertical scaling (changing tier) and horizontal scaling via Redis Cluster (enabling clustering). For HA, they can create a standard tier instance with replication (a read replica).

Answer

Use Cloud Storage snapshots for persistence

Answer

Enable AOF persistence

Question 17

A developer wants to manually promote a release from a staging target to a production target using Cloud Deploy. Which gcloud command should they use?

Accepted Answer

gcloud deploy releases promote. The `gcloud deploy releases promote` command is the correct choice because Cloud Deploy uses a promotion-based model where releases are advanced through targets (e.g., staging to production) via a promote action. This command triggers the creation of a new rollout in the next target in the promotion sequence, effectively moving the release forward without manual rollout creation.

Answer

gcloud deploy releases approve

Answer

gcloud deploy rollouts create

Answer

gcloud deploy targets promote

Question 18

A team uses Cloud Build to deploy a Python service to Cloud Run. They need to ensure the service uses a custom domain and only accepts HTTPS traffic. Which flags should they include in the gcloud run deploy command?

Accepted Answer

No flags related to custom domain; use gcloud beta run domain-mappings create separately. The `gcloud run deploy` command does not include flags for mapping a custom domain. Custom domain mapping is a separate step that must be performed using the `gcloud beta run domain-mappings create` command (or via the Cloud Run console). The `--ingress` flag controls traffic routing (e.g., all, internal, internal-and-cloud-load-balancing), not domain configuration, and HTTPS is enforced by default on Cloud Run services.

Answer

--ingress=internal --platform=managed

Answer

--ingress=all --allow-unauthenticated

Answer

--ingress=all --no-allow-unauthenticated

Question 19

A financial institution uses BigQuery for BI reporting. They have a table 'transactions' (10 TB) partitioned by transaction_date and clustered by customer_id. A common report filters on customer_id and last 30 days. The report is slow. Which change would most improve query performance for this specific report?

Accepted Answer

Add clustering on transaction_date in addition to customer_id. The report filters on customer_id and the last 30 days. The table is partitioned by transaction_date and clustered by customer_id. Adding clustering on transaction_date (in addition to customer_id) allows BigQuery to sort rows within each partition by transaction_date, enabling block pruning for the date range filter in addition to partition pruning. This reduces the amount of data scanned and improves query performance. Option A is incorrect because changing the partition column to customer_id would not help with the date filter, which is a key part of the report. Option B is wrong because removing clustering would increase scanned data for the customer_id filter. Option D is unnecessary because BigQuery automatically manages clustering metadata; manual reclustering is not required.

Answer

Change partition column to customer_id

Answer

Remove clustering and rely only on partitioning

Answer

Manually recluster the table daily

Question 20

An organization requires a disaster recovery solution for Cloud SQL with a target RPO of less than 5 minutes and RTO of less than 2 hours. The primary region is us-central1, and the DR region is us-west1. Which approach should they use?

Accepted Answer

Create a cross-region read replica and promote it during disaster. Cross-region read replicas provide an RPO equal to replication lag (typically <5s) and RTO of minutes (manual promotion). This meets the requirements. Cross-region backup restore would have an RPO equal to the last backup age (hours) and RTO of hours. HA failover is for same-region only. PITR alone does not provide cross-region DR.

Answer

Use point-in-time recovery with daily backups and restore into a new instance

Answer

Configure a scheduled export of the database to Cloud Storage and import in DR region

Answer

Enable HA configuration in a single region