PCDE Practice Test 22 — 15 Questions

Question 1

An application running on Cloud Run is experiencing high tail latency. The team wants to visualize which functions are consuming the most CPU time. Which Google Cloud tool should they use and how should they instrument the application?

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

Accepted Answer

Cloud Profiler with OpenTelemetry or Profiler agent. Cloud Profiler is Google Cloud's continuous CPU and heap profiling service, and it is the correct tool to identify which functions consume the most CPU time. It is instrumented via the Profiler agent (or OpenTelemetry for supported runtimes) and produces flame graphs showing per-function CPU usage, which directly addresses the tail-latency investigation.

Answer

Cloud Monitoring with custom metrics for CPU

Answer

Cloud Trace with automatic instrumentation

Answer

Cloud Logging with structured logs

Question 2

A startup is building an IoT analytics platform that ingests sensor data at high velocity and needs to run real-time dashboards and ad-hoc queries on the data. Which TWO Google Cloud databases should they use together? (Choose 2)

Accepted Answer

Cloud Bigtable. Cloud Bigtable (A) is correct because it is a fully managed, horizontally scalable NoSQL wide-column store designed for high-throughput, low-latency ingestion of time-series and sensor data at very high velocity, making it ideal for the IoT ingest layer. BigQuery (C) is correct because it is a serverless, columnar analytics data warehouse that supports real-time streaming inserts and enables fast ad-hoc SQL queries and dashboarding over large datasets. Together they form a common pattern: Bigtable absorbs the high-velocity writes, while BigQuery provides the analytical query engine for dashboards and exploration. Cloud Spanner (B) is a globally distributed relational database optimized for transactional consistency, not high-velocity IoT ingestion plus analytics. Firestore (D) is a document database for mobile/web app data with lower throughput characteristics, and Cloud SQL (E) is a managed relational database that does not scale to IoT-scale ingest or analytical workloads.

Answer

Cloud Spanner

Answer

Firestore

Answer

Cloud SQL

Question 3

An organization needs to continuously replicate data from an on-premises PostgreSQL database to AlloyDB with minimal downtime during cutover. They have set up DMS continuous migration. During the cutover phase, what is the correct sequence of steps?

Accepted Answer

Quiesce writes to source, confirm DMS lag is 0, promote destination, update connection strings, verify application, keep source running read-only.. The standard cutover: quiesce writes, verify DMS lag is 0, promote (stop replication), update connection strings, verify app, and keep source for rollback.

Answer

Update connection strings first, then promote destination, then verify lag.

Answer

Stop source database, promote destination, then update connection strings.

Answer

Promote target immediately, then verify lag, then update connection strings.

Question 4

A company is deploying a microservice on Cloud Run that needs to handle up to 1000 concurrent requests per instance. The default concurrency setting is 80. How should they configure the service to achieve the desired concurrency?

Accepted Answer

Set the `--concurrency` flag to 1000.. Cloud Run allows setting max concurrent requests per instance via the `--concurrency` flag. Setting it to 1000 enables the desired throughput.

Answer

Set the `--execution-environment` to gen2.

Answer

Set the `--cpu` flag to 4.

Answer

Set the `--max-instances` flag to 1000.

Question 5

During a cutover from an on-premises Oracle database to Cloud SQL for PostgreSQL, the team needs to minimize downtime. They have set up continuous replication using a custom tool. Which sequence of steps should they follow to perform the cutover with minimal data loss?

Accepted Answer

Quiesce writes, verify lag is 0, promote destination, update connection strings, verify application.. The correct cutover sequence for minimal data loss with continuous replication is: first quiesce writes to the source to stop new transactions, then verify replication lag is zero to ensure all data has been applied to the destination, then promote the destination to primary, update application connection strings, and finally verify the application. This order guarantees no data loss and a clean transition.

Answer

Quiesce writes, promote destination, verify lag, update connection strings.

Answer

Promote destination, quiesce writes, verify lag, update connection strings, verify application.

Answer

Update connection strings, quiesce writes, verify lag, promote destination.

Question 6

You are configuring an alerting policy in Cloud Monitoring to notify when the 99th percentile latency of your application exceeds 500 ms for 5 minutes. Which metric type and reducer should you use?

Accepted Answer

DISTRIBUTION metric type with PERCENTILE reducer. The 99th percentile latency is a statistical measure that requires examining the full distribution of latency samples, not just a single aggregate value. Cloud Monitoring represents such multi-sample data as a DISTRIBUTION metric type, and the PERCENTILE reducer is specifically designed to extract a given percentile (e.g., 99th) from that distribution. Combining DISTRIBUTION with PERCENTILE lets the alerting policy evaluate the p99 value against the 500 ms threshold over the 5-minute window.

Answer

CUMULATIVE metric type with MEAN reducer

Answer

GAUGE metric type with MAX reducer

Answer

DELTA metric type with SUM reducer

Question 7

A database engineer is designing a data model for a BI dashboard that tracks daily sales by product category. The data source is a transactional database with a normalized schema. Which BigQuery feature should they use to update the fact table incrementally each day?

Accepted Answer

Scheduled queries with MERGE statements. Scheduled queries with MERGE statements allow incremental updates by inserting new rows and updating existing ones based on a unique key, such as date and product category. This avoids full table reloads, making it efficient for daily fact table refreshes from a normalized transactional source.

Answer

Streaming inserts

Answer

BigQuery Data Transfer Service

Answer

Load jobs with WRITE_TRUNCATE

Question 8

A company has a Cloud SQL for MySQL instance with point-in-time recovery (PITR) enabled. They need to restore the database to a specific time exactly 2 hours ago to recover from an accidental data deletion. What is the minimum requirement for this operation?

Accepted Answer

Binary logging must be enabled, and the transaction log retention period must cover the desired restore time. Cloud SQL PITR relies on binary logging (binary log for MySQL) to replay transactions from the last backup to the desired point in time. For a restore to 2 hours ago, binary logging must be enabled and the transaction log retention must cover that 2-hour window. Without sufficient log retention, PITR cannot reach the target time.

Answer

The instance must have a backup window configured within the last 2 hours

Answer

The instance must have at least one automatic backup taken after the desired restore time

Answer

The instance must be stopped before initiating the restore

Question 9

Refer to the exhibit. The company plans to store 3 TB of data in this instance. What is the minimum number of nodes required? (Assume 2 TB per node for HDD and 4 TB per node for SSD; this instance uses SSD.)

Accepted Answer

1. The question explicitly assumes 4 TB per SSD node and asks for the minimum number of nodes required to store 3 TB of data. Since 4 TB is greater than 3 TB, a single node provides sufficient raw storage capacity. The question does not specify any requirements for high availability, replication, or redundancy, so the minimum number of nodes is 1.

Answer

4

Answer

2

Answer

3

Question 10

A development team uses Cloud Spanner and needs to add a new index to a table with 10 TB of data. They require that the operation does not block writes or cause downtime. Which approach should they take?

Accepted Answer

Execute 'CREATE INDEX' statement; Spanner will build the index in the background without blocking.. Cloud Spanner supports online schema changes, including CREATE INDEX, which build in the background without blocking reads or writes to the table. This is a core differentiator of Spanner's distributed architecture, allowing index creation on large tables without downtime. The operation is transparent to applications and does not require a maintenance window.

Answer

Create a new table with the index and batch copy data from the original table.

Answer

Schedule a maintenance window, then create the index during the window.

Answer

Drop the table and recreate it with the index, then restore from backup.

Question 11

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. A cross-region read replica is correct because it continuously replicates data from the primary region to the DR region, providing an RPO of seconds (well under 5 minutes). During a disaster, the replica can be promoted to a standalone instance, achieving an RTO of minutes to under 2 hours. This is a standard DR pattern for Cloud SQL.

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

Question 12

You are designing a schema for a Cloud SQL for PostgreSQL database that supports full-text search across millions of product descriptions. The application requires fast search results ranked by relevance. Which schema design is most appropriate?

Accepted Answer

Use a tsvector column with a GIN index on that column. PostgreSQL's tsvector data type, combined with a GIN index, is specifically designed for full-text search. It preprocesses text into lexemes, supports stemming and ranking, and the GIN index enables fast lookups for millions of rows, meeting the requirement for relevance-ranked results.

Answer

Use a separate Elasticsearch instance

Answer

Use a LIKE '%term%' query with a B-tree index

Answer

Use materialized view with trigram indexes

Question 13

A company runs an e-commerce platform on Cloud Spanner multi-region (nam6). They experience a regional failure affecting the leader region. After failover, they observe increased write latency. What is the most likely cause?

Accepted Answer

The new leader region is geographically farther from the application instances.. In Cloud Spanner multi-region configurations, the leader region handles all writes, and the other regions contain read-write replicas that must acknowledge writes via quorum. When a regional failure forces a new leader to be elected in a different region, writes must now traverse the distance between the application and the new leader region, and quorum acknowledgment latency increases because the new leader may be farther from some replicas and from the application. This geographic distance is the most likely cause of increased write latency.

Answer

The new leader region has fewer replicas.

Answer

The instance is now running in read-only mode.

Answer

The failover caused data loss, requiring recovery.

Question 14

A site reliability team wants to define an SLO for a service with a target availability of 99.9% over a 30-day window. The error budget is exhausted. Which action MUST the team take?

Accepted Answer

Freeze all non-critical releases until the budget recovers.. When an SLO's error budget is exhausted, the standard SRE practice is to freeze non-critical releases and focus engineering effort on reliability improvements until the service is back within its error budget. Continuing normal feature development while the budget is exhausted violates the core purpose of the error budget policy, which is to enforce a balance between innovation and reliability.

Answer

Ignore the budget and continue normal development.

Answer

Deploy a new feature to attract more users.

Answer

Increase the SLO to 99.95% to make up for lost budget.

Question 15

A company uses Firestore in Native mode and has a collection with many documents containing array fields. They want to index the array values to support queries like 'array-contains'. What is the correct approach?

Accepted Answer

No additional index configuration is required; arrays are automatically indexed. Firestore in Native mode automatically indexes all fields within documents, including array fields, so queries using 'array-contains' work without any manual index creation. Single-field indexes are created by default for every field in a collection, which is why no additional configuration is needed for basic array queries. This automatic indexing is a core design feature of Firestore that differentiates it from Datastore mode.

Answer

Create an index exemption for the array field to enable indexing

Answer

Switch to Datastore mode and configure indexes manually

Answer

Create a composite index that includes the array field