PCDE Manage a solution that can span multiple database technologies Practice Test 6 — 15 Questions

Question 1

A company is using Cloud Spanner and wants to back up a database to Cloud Storage for long-term retention. They also need to restore the database to a specific point in time within the last 7 days. Which two features should they use? (Choose TWO.)

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

Accepted Answer

Use Spanner's point-in-time recovery feature to restore to a specific timestamp. Option B is correct because Cloud Spanner's point-in-time recovery (PITR) lets you restore a database to any timestamp within the configurable retention window (up to 7 days by default, extendable to 30 days), which exactly matches the requirement to restore to a specific point in time within the last 7 days. Option E is correct because `gcloud spanner backups create` creates a Spanner backup that can be stored and retained for long-term retention, and backups can also be exported to Cloud Storage for archival purposes. Option A is incorrect because a cross-region replica is for availability and read scaling, not for backup or point-in-time restore. Option C is incorrect because version management is an object-versioning concept (e.g., Cloud Storage), not a Spanner instance feature. Option D is incorrect because `gcloud spanner databases export` is not the mechanism for creating a restorable backup; Spanner backups are created with `gcloud spanner backups create`.

Answer

Set up a cross-region replica for backup

Answer

Enable version management on the Spanner instance

Answer

Export the database using gcloud spanner databases export

Question 2

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. Option B is correct because enabling Redis Cluster mode on a Memorystore for Redis instance shards data across multiple nodes, which is exactly what is needed to distribute data and scale horizontally for increased traffic. Option C is correct because upgrading the instance to a higher memory size by changing the tier increases capacity, allowing the instance to handle more data and traffic while maintaining performance. Option D is correct because creating a read replica in a different zone provides high availability and read scalability, ensuring the instance remains available if the primary zone fails. Option A is incorrect because Cloud Storage snapshots are not a Memorystore for Redis persistence or scaling mechanism; Redis persistence uses RDB snapshots managed by the service, not Cloud Storage. Option E is incorrect because enabling AOF persistence improves durability by logging every write operation, but it does not provide high availability or data distribution across nodes, and Memorystore for Redis does not expose AOF as a scaling action.

Answer

Use Cloud Storage snapshots for persistence

Answer

Enable AOF persistence

Question 3

A company is using Bigtable for a high-throughput write workload. They need to monitor replication lag between clusters in a replicated setup and ensure that reads are eventually consistent. Which two configurations should they check? (Choose TWO.)

Accepted Answer

Ensure replication is set to asynchronous (default). Option C is correct because Bigtable replication is asynchronous by default, and this asynchronous model is what allows writes to propagate to other clusters with eventual consistency; verifying that replication is indeed set to asynchronous (the default) confirms the expected behavior for the replicated setup. Option D is correct because the Cloud Monitoring metric 'bigtable.googleapis.com/cluster/replication_lag' is the specific metric that reports replication lag between clusters, so monitoring it directly addresses the requirement to track lag. Option A is not correct because a Data Studio dashboard is only a visualization layer and does not itself configure or validate replication lag or consistency. Option B is not correct because Bigtable does not expose a configurable 'strong' replication consistency level for cross-cluster replication. Option E is not correct because Bigtable does not support synchronous replication with zero lag; replication is asynchronous.

Answer

Create a replication dashboard in Data Studio

Answer

Set the replication consistency level to 'strong'

Answer

Enable synchronous replication for zero lag

Question 4

A media analytics firm ingests clickstream events into Cloud Bigtable and simultaneously needs low‑latency point lookups for user profiles from a Cloud SQL for PostgreSQL database. The team wants a single application layer that can join these datasets without duplicating data into Cloud SQL. Which Google Cloud service should they use to query both Bigtable and Cloud SQL from one SQL statement?

Accepted Answer

BigQuery federated queries with external tables over Cloud Bigtable and Cloud SQL. BigQuery federated queries allow a single SQL statement to read from Cloud Bigtable through the Bigtable external data source and from Cloud SQL through the Cloud SQL connector. This provides a unified query layer without copying data, satisfying the need to join clickstream events with user profiles directly from both databases.

Answer

Cloud Spanner with foreign data wrappers to Cloud Bigtable and Cloud SQL

Answer

Cloud Data Fusion with a Bigtable source and a Cloud SQL sink to materialize the join

Answer

Dataproc with Hive external tables pointing to Cloud Bigtable and Cloud SQL via JDBC

Question 5

A multinational retail company operates a Cloud Spanner instance for its order management system in the us-central1 region. Due to data sovereignty requirements, they need to store customer data in the europe-west1 region as well, while maintaining strong consistency for global transactions. They also need to ensure that reads from europe-west1 do not incur cross-region latency for local users. Which Cloud Spanner configuration should they implement?

Accepted Answer

Add a read-only replica in europe-west1 to the existing Cloud Spanner instance.. Adding a read-only replica in europe-west1 to the existing Cloud Spanner instance enables local reads in Europe with low latency while preserving strong consistency for writes. This approach meets data sovereignty requirements by storing data in the European region and avoids the complexity of managing separate instances or non-native replication methods. It is the standard way to extend read capabilities across regions in Cloud Spanner.

Answer

Use Cloud Spanner's change streams to replicate data to a Cloud Bigtable instance in europe-west1 for local reads.

Answer

Create a new Cloud Spanner instance in europe-west1 and use Cloud Spanner replication to synchronize data between the two instances.

Answer

Configure the existing Cloud Spanner instance to use a multi-region configuration that includes both us-central1 and europe-west1.

Question 6

A media analytics company uses Cloud Bigtable for time-series sensor data. They need to run a complex analytical query that joins Bigtable data with a Cloud SQL for PostgreSQL table containing device metadata. The query must run without exporting data, and the results must be available within minutes. Which approach should they use?

Accepted Answer

Use BigQuery federated queries to join Bigtable and Cloud SQL data.. BigQuery federated queries allow querying data directly from Cloud Bigtable and Cloud SQL without exporting. This enables complex joins and analytics with fast results, satisfying the no-export and minute-level latency requirements. Other options either lack SQL join capability, require data movement, or involve custom pipeline development.

Answer

Use Cloud Bigtable's SQL interface to directly join with Cloud SQL.

Answer

Export Bigtable data to Cloud Storage and load into Cloud SQL for a join.

Answer

Use Dataflow to read from both sources and perform the join in a pipeline.

Question 7

A financial analytics platform stores tick data in Cloud Bigtable and reference data in Cloud SQL for PostgreSQL. Analysts need near-real-time joins between the two datasets using standard SQL, and the solution must minimize operational overhead. What should the team do?

Accepted Answer

Create a federated query in BigQuery that reads Cloud Bigtable and Cloud SQL for PostgreSQL, and schedule refreshes of the reference data.. BigQuery federated queries are designed to query external data sources including Cloud Bigtable and Cloud SQL for PostgreSQL using standard SQL. By using federated queries, analysts can join tick data with reference data without building and operating a separate replication pipeline. This directly addresses the near-real-time analysis and low operational overhead goals.

Answer

Enable Cloud SQL for PostgreSQL foreign data wrappers to query Cloud Bigtable directly.

Answer

Configure Cloud Spanner as a mid-tier and replicate both datasets into it using Dataflow.

Answer

Export Cloud Bigtable to Cloud Storage and load it into Cloud SQL for PostgreSQL nightly, then join locally.

Question 8

A financial services company uses Cloud SQL for PostgreSQL to store transaction data. They are required to maintain a real-time copy of the database in a different region for disaster recovery and analytics. The analytics team needs to run complex queries without impacting the production database. The company wants to minimize data loss in case of a regional failure and ensure the replica is always available. Which two actions should they take? (Choose two.)

Accepted Answer

Set up a Cloud SQL replica in a different region using the gcloud sql instances create --replica-type=READ command.. To achieve a real-time cross-region copy for disaster recovery and analytics, you must create a cross-region read replica. This involves configuring the replica using the appropriate gcloud command or console. Enabling binary logging is a prerequisite for replication, but the two key actions are creating the replica and ensuring it is in a different region. High availability and point-in-time recovery do not provide the required continuous replication and offloading capabilities.

Answer

Use Cloud SQL's point-in-time recovery to restore the database to a different region.

Answer

Enable binary logging on the primary instance.

Answer

Enable high availability (HA) on the primary instance.

Question 9

A media company stores time-series clickstream events in Cloud Bigtable. Analysts run ad hoc scans over the last 90 days, and the table has grown to hundreds of terabytes. The team wants to reduce storage cost for data older than 90 days while keeping the most recent data on fast storage. They also want queries to remain efficient. What should they do?

Accepted Answer

Create a Cloud Bigtable table with a tiered storage policy that moves data older than 90 days to a colder tier. Cloud Bigtable tiered storage lets you define an age-based policy so that data older than a threshold moves to a lower-cost tier automatically. Because the data stays in the same table, ad hoc scans across the 90-day window continue to work, while the bulk of hundreds of terabytes that is older than 90 days no longer pays the highest storage rate.

Answer

Configure a Cloud Bigtable table with a change stream and read older data through the stream

Answer

Create a Cloud Storage bucket and export data older than 90 days with a Cloud Bigtable export job

Answer

Use Cloud Bigtable garbage collection policies to delete data older than 90 days

Question 10

Your organization runs a global supply-chain application whose transactional data lives in Cloud Spanner. A new analytics team must query this data using standard SQL joins and aggregations with sub-second latency, but they have no Spanner expertise and want to avoid impacting transactional workloads. Which approach should you recommend?

Accepted Answer

Export the Spanner database to Cloud Storage using Dataflow and load it into BigQuery, then let analysts query BigQuery.. BigQuery is Google Cloud's serverless analytics warehouse, and the managed Dataflow export/import path is the recommended way to mirror Cloud Spanner data into it. This decouples analytical query load from the transactional instance, gives analysts standard SQL, and delivers the sub-second interactive performance described. The other choices either keep queries on Spanner's engine or provide change data rather than a queryable analytics store.

Answer

Enable the Spanner PostgreSQL interface on the existing instance so analysts can use standard PostgreSQL tooling.

Answer

Create a Cloud Spanner change stream and have analysts subscribe to it with a custom application.

Answer

Create a Cloud Spanner read-only replica in a nearby region and give the analysts direct access to it.

Question 11

A startup uses Cloud Firestore for its mobile app's user profiles and Cloud SQL for PostgreSQL for billing data. They need to generate a daily report that combines user profile information with billing records. The report must be generated with minimal operational overhead. What should they do?

Accepted Answer

Use BigQuery to query Firestore and Cloud SQL via federated queries.. BigQuery federated queries can directly query Cloud SQL and Firestore data, allowing the startup to join and generate reports without data movement. BigQuery is serverless, so operational overhead is minimal. This approach is efficient for daily reporting and leverages existing data sources.

Answer

Export Firestore data to Cloud Storage and load into Cloud SQL, then run a SQL join.

Answer

Use Cloud Dataflow to read from both Firestore and Cloud SQL, join the data, and write to a report.

Answer

Use Cloud Composer to orchestrate a workflow that extracts, transforms, and loads the data.

Question 12

A retail company runs its transactional workload on Cloud SQL for MySQL and its product catalog on Firestore in Native mode. The mobile app needs to read both the user's order history and the product catalog in one request. Which approach best supports this?

Accepted Answer

Build an API service that calls the Cloud SQL Admin API and Firestore REST API, then returns a combined response.. When an application uses multiple database technologies, a common pattern is to place a backend service between the client and the databases. The service can query Cloud SQL for order history and Firestore for the catalog using each service's native API, then assemble one response. This preserves the strengths of each database and gives the mobile app a single request.

Answer

Migrate the product catalog into Cloud SQL for MySQL so both datasets are in one database.

Answer

Create a Cloud SQL federated table that points to Firestore collections.

Answer

Enable the Firestore in Datastore mode API and query Cloud SQL through the same SDK.

Question 13

A media analytics company stores clickstream events in Cloud Bigtable and serves low-latency user-profile reads from Cloud Spanner. A nightly Spark job must join recent Bigtable events with Spanner profile rows to produce an aggregated dataset. The team wants a managed, serverless execution path that reads both stores without provisioning a persistent cluster or copying data between them. Which approach should they choose?

Accepted Answer

Run the join as a Dataproc Serverless for Spark batch workload, using the Cloud Bigtable HBase client and the Spanner Spark connector in the same job.. Dataproc Serverless for Spark is the managed, serverless batch option that supports both the Cloud Bigtable HBase client and the Spanner Spark connector in one job, so the join can read both databases directly without copying data. Exporting to Cloud Storage and loading into Spanner adds latency and duplication, while BigQuery federated queries and a persistent Dataproc cluster do not fit the stated constraints.

Answer

Deploy a long-running Dataproc on Compute Engine cluster with Bigtable and Spanner connectors, and schedule the join with Cloud Scheduler.

Answer

Create a federated query in BigQuery that reads Cloud Bigtable through the Bigtable external table connector and Cloud Spanner through the Spanner external dataset.

Answer

Export Cloud Bigtable rows to Cloud Storage using a Dataflow template, then load them into Cloud Spanner and run the join with Spanner SQL.

Question 14

A healthcare company stores patient records in Cloud Spanner. They need to run analytical queries that join Spanner data with data in Cloud Bigtable for real-time anomaly detection. The queries require low-latency reads and must not impact the performance of the transactional workload. Which approach should they use?

Accepted Answer

Use BigQuery to query both Spanner and Bigtable via federated queries, then join the results. BigQuery federated queries can access both Cloud Spanner and Cloud Bigtable, enabling analytical joins without affecting the transactional Spanner workload. This separates concerns: Spanner handles OLTP, Bigtable serves low-latency NoSQL access, and BigQuery performs the analytical join. The other options either do not support the required integration, introduce data duplication and latency, or add load to Spanner.

Answer

Create a Cloud Dataflow pipeline that streams Bigtable data into Spanner in real time

Answer

Export Bigtable data to Cloud Storage and load it into Spanner as a separate table

Answer

Use Cloud Spanner federated queries to query Bigtable directly from Spanner

Question 15

A healthcare company uses Cloud Bigtable to store patient vitals and Cloud Spanner for patient records. They need to run a query that finds all patients with a specific condition and retrieves their latest vitals. The query must be executed with minimal latency and should not require copying data. Which solution meets these requirements?

Accepted Answer

Use BigQuery federated queries to join Cloud Spanner and Cloud Bigtable data.. BigQuery federated queries allow querying Cloud Spanner and Cloud Bigtable directly, enabling joins without data copying. This provides low-latency results and meets the healthcare company's requirements. Other options involve data movement, unsupported features, or additional pipeline development, which are not suitable.

Answer

Use Cloud Dataflow to read from both sources, perform the join, and write results to a cache.

Answer

Export Cloud Spanner data to Cloud Storage and load into Bigtable, then query Bigtable.

Answer

Use Cloud Spanner's SQL to query Bigtable via a foreign data wrapper.