PCDE Practice Test 42 — 15 Questions

Question 1

An organization wants to deploy a Cloud Run service using Cloud Deploy. They need to run a database migration script before each new revision starts serving traffic. Which Cloud Deploy feature should they use?

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

Accepted Answer

Configure a preDeploy hook that runs a Cloud Run Job. Cloud Deploy supports preDeploy and postDeploy hooks that run as Cloud Run Jobs (or other execution environments) at specific points in the rollout. A preDeploy hook runs before the new revision receives traffic, making it the correct place for a database migration that must complete before the new code serves requests. This ensures schema changes are applied before the new revision starts handling traffic.

Answer

Use a Cloud Build step in the delivery pipeline

Answer

Add a manual approval gate before deployment

Answer

Use a postDeploy hook to run the migration after traffic is switched

Question 2

Which of the following is a benefit of using parent-child interleaved tables in Cloud Spanner?

Accepted Answer

Improved read performance for queries joining parent and child. Parent-child interleaved tables in Cloud Spanner physically co-locate child rows with their parent row on the same split, so a join between parent and child can be satisfied by a single local lookup rather than a distributed lookup across splits. This locality dramatically reduces latency and network hops for parent-child join queries. Interleaving is therefore recommended specifically to improve read performance for hierarchical data access patterns.

Answer

Increased write throughput by distributing writes

Answer

Automatic sharding across regions

Answer

Eliminates the need for secondary indexes

Question 3

An organization is implementing Binary Authorization for GKE. They need to ensure that only container images signed by their CI system are deployed. Which service must be enabled and configured to enforce this?

Accepted Answer

Binary Authorization attestor and policy. Binary Authorization for GKE enforces deploy-time verification of container images against attestations. To enforce that only images signed by the CI system are deployed, you must configure a Binary Authorization attestor (which holds the public key used to verify attestations) and a Binary Authorization policy (which defines the rules requiring attestations for specific clusters/namespaces). These two components together are what enforce the admission control decision.

Answer

Cloud Key Management Service (KMS)

Answer

Artifact Registry with Container Analysis

Answer

Cloud Build service account with 'iam.serviceAccountUser' role

Question 4

An SRE team defines an SLO for a batch processing pipeline. Which SLI is most appropriate for pipeline freshness?

Accepted Answer

Age of the most recent output record. Pipeline freshness measures how up-to-date the output data is, which is best expressed as the age of the most recent output record (or the time since the last successful batch produced data). This SLI directly captures staleness, which is what a freshness SLO is intended to bound.

Answer

Percentage of records with correct values

Answer

Number of records processed per second

Answer

Number of failed pipeline runs

Question 5

An SRE team wants to reduce toil by automating the response to common alert notifications. For example, when a disk usage alert fires, they want to automatically run a script to clean up temporary files. Which Google Cloud service is best suited for this?

Accepted Answer

Cloud Functions. Cloud Functions is Google Cloud's serverless, event-driven compute service, ideal for running small scripts in response to events such as Pub/Sub alert messages. When a disk usage alert fires, it can publish to a Pub/Sub topic that triggers a Cloud Function to run the cleanup script. This directly addresses the SRE goal of automating toil with minimal infrastructure.

Answer

Cloud Build

Answer

Workflows

Answer

Cloud Scheduler

Question 6

A financial services company runs a high-frequency trading application that requires strong consistency, horizontal scalability, and low-latency transactions across multiple regions. Which Google Cloud database should they choose?

Accepted Answer

Cloud Spanner. Cloud Spanner is a globally distributed, strongly consistent, horizontally scalable relational database service designed for mission-critical applications like trading. Cloud SQL is not multi-region, Bigtable does not support SQL/ACID, and Firestore is not relational.

Answer

Cloud SQL

Answer

Cloud Bigtable

Answer

Firestore

Question 7

You are designing a Cloud SQL for PostgreSQL database. The application has a table with 1 million rows that is frequently queried using equality on the 'email' column and range queries on the 'created_at' column. Which index strategy minimizes query latency?

Accepted Answer

Create a composite B-tree index on (email, created_at).. A composite B-tree index on (email, created_at) allows the database to satisfy both the equality condition on 'email' and the range condition on 'created_at' in a single index scan. PostgreSQL can use the leftmost column for equality filtering and then traverse the index tree to retrieve the range portion efficiently, minimizing random I/O and query latency.

Answer

Create a full-text index on email.

Answer

Create a B-tree index on email only.

Answer

Create separate B-tree indexes on email and created_at.

Question 8

A DevOps team is using Pub/Sub to process high-volume event streams. They notice that the subscriber is falling behind and messages are being redelivered frequently. They need to increase throughput. Which TWO actions should they take? (Choose TWO)

Accepted Answer

Increase the acknowledgement deadline to give subscribers more time to process messages. Option A is correct because extending the acknowledgement deadline prevents Pub/Sub from redelivering messages while a subscriber is still processing them; if the deadline expires before ack, the message is redelivered, so a longer deadline reduces duplicate deliveries and lets slow subscribers finish work, improving effective throughput. Option D is correct because Pub/Sub pull subscriptions scale horizontally: adding more parallel pull consumers (threads/processes/clients) increases the number of concurrent StreamingPull streams and outstanding message leases, raising overall message processing throughput. Option B is not appropriate here because message ordering keys enforce per-key ordered delivery and can serialize processing, which typically reduces throughput rather than increasing it. Option C is incorrect because switching from pull to push does not inherently deliver messages faster; push and pull have similar delivery characteristics, and push adds endpoint/retry considerations without solving subscriber lag. Option E is incorrect because decreasing flow control max outstanding messages lowers the number of messages a subscriber can hold in flight, which throttles throughput and would worsen the backlog.

Answer

Enable message ordering keys to ensure orderly processing

Answer

Change the subscription type from pull to push for faster delivery

Answer

Decrease the flow control max outstanding messages to reduce load on subscribers

Question 9

A team is migrating a 2 TB MySQL database from on-premises to Cloud SQL. They want to minimize downtime. The source is MySQL 8.0 with InnoDB tables, and the application can be read-only during cutover. Which approach provides the lowest downtime while ensuring data consistency?

Accepted Answer

Use Database Migration Service with continuous CDC and promote when lag is zero.. DMS with continuous CDC provides near-zero downtime by replicating ongoing changes. After the initial dump, the source continues to replicate changes, allowing a quick cutover with minimal downtime.

Answer

Use mysqldump with --single-transaction to export, then import to Cloud SQL using mysql client.

Answer

Use Cloud Dataflow to stream data from MySQL to Cloud SQL.

Answer

Use gcloud sql import command with a compressed dump file from Cloud Storage.

Question 10

An organization uses Cloud SQL for MySQL and needs to perform disaster recovery testing by failing over to a cross-region read replica without impacting the primary instance. They want to validate the promotion process and measure the actual RTO. Which approach should be used?

Accepted Answer

Promote a cross-region read replica in an isolated project or non-production environment. Promoting a cross-region read replica in an isolated project or non-production environment lets the team validate the promotion process and measure RTO without touching the production primary. This is the standard DR-testing pattern because promotion is a destructive, one-way operation on the replica. Isolating the test avoids any risk to production traffic or replication topology.

Answer

Create a clone of the primary instance in the target region and test failover

Answer

Use Cloud SQL’s switchover feature for HA instances

Answer

Run a gcloud command to promote the existing read replica in the production project

Question 11

An e-commerce platform runs on Cloud SQL and expects heavy write traffic during a flash sale. The database instance currently has 8 vCPUs and 32 GB RAM. Based on the max_connections formula (max_connections = RAM_MB/16), what is the current maximum number of connections, and what change would increase it?

Accepted Answer

2048 connections; increase RAM to 64 GB. Using the formula max_connections = RAM_MB/16, 32 GB RAM equals 32,768 MB, so 32,768/16 = 2,048 connections. To increase max_connections, RAM must be increased—doubling RAM to 64 GB yields 4,096 connections. vCPU changes do not affect this formula.

Answer

2048 connections; increase vCPU to 16

Answer

4096 connections; increase vCPU to 16

Answer

4096 connections; increase RAM to 64 GB

Question 12

An organization has a Cloud Deploy delivery pipeline with a canary deployment strategy to GKE. They want to automatically pause the rollout if the canary revision's error rate exceeds 5% for 5 minutes. Which feature should they use?

Accepted Answer

Canary deployment strategy with metric analysis. Cloud Deploy's canary deployment strategy supports metric analysis, which lets you define a success threshold (e.g., error rate < 5%) and a duration (e.g., 5 minutes). Cloud Deploy evaluates the canary revision against those metrics and automatically pauses or rolls back the rollout if the threshold is breached, without human intervention. This is exactly the automated, metric-driven pause the scenario requires.

Answer

PreDeploy hook

Answer

Rollback on deploy failure

Answer

Manual approval gate

Question 13

A DevOps engineer wants to create an alerting policy that fires when the 99th percentile of request latency exceeds 2000 ms for any 5-minute window. Which metric type and aggregation should they use?

Accepted Answer

Use a distribution metric with reducer 99th percentile. A distribution metric with a 99th percentile reducer is correct because distribution metrics capture the full statistical spread of latency values, allowing percentile aggregations like p99. A 5-minute alignment window with a p99 reducer produces exactly the '99th percentile over any 5-minute window' condition the engineer needs.

Answer

Use a GAUGE metric with reducer MAX

Answer

Use a CUMULATIVE metric with reducer DELTA

Answer

Use a DELTA metric with reducer COUNT

Question 14

A media analytics company runs a Cloud Bigtable instance with a single cluster in us-central1. The application performs sequential read-modify-write operations on individual rows, and latency spikes occur during peak ingest. You must reduce hot-spotting on individual tablet servers while keeping the row key design schema-free. What should you do?

Accepted Answer

Reverse the byte order of the row key or hash a prefix so writes spread across the key space.. Cloud Bigtable distributes rows by lexicographic key ranges, so a monotonically increasing or narrow key concentrates writes on one tablet server. Reversing the key or adding a hashed prefix spreads those writes across the key space, letting many tablet servers share the load. Node count and garbage collection policies do not change key distribution, and replication addresses read scaling and durability rather than localized write hot spots.

Answer

Enable a column family garbage collection policy that drops cells older than one day.

Answer

Configure a replication cluster in a second region and route writes to both clusters.

Answer

Increase the number of nodes in the cluster to add more tablet servers.

Question 15

An e-commerce application uses Cloud Spanner for its global inventory database. The application experiences high write latency during peak hours. After reviewing the schema, the database engineer notices that the primary key is an auto-incrementing integer. What is the most likely cause of the high write latency?

Accepted Answer

The monotonically increasing primary key creates a hotspot.. Cloud Spanner distributes data across splits based on the primary key range. A monotonically increasing integer primary key, such as an auto-incrementing ID, causes all new writes to target the same split (the highest key range), creating a hotspot. This single split becomes a bottleneck, leading to high write latency during peak hours, as Spanner cannot parallelize the writes across multiple nodes.

Answer

The application is using read-only transactions instead of read-write transactions.

Answer

The database is configured with a backup retention period of 2 seconds.

Answer

The application is using read replicas that are out of date.