1. You are designing a BigQuery database with multiple tables in a single dataset. The data stored in the dataset is measurement data from sensors on vehicles in the company’s fleet. Data is collected on each vehicle and downloaded at the end of each shift. After that, it is loaded into a partitioned table. You want to have efficient access to the most interesting data, which you define as a particular measurement having a value greater than 100.00. You want to cluster on that measurement column, which is a FLOAT64. When you define the table with a timestamped partitioned table and clustering on the measurement column, you receive an error. What could that error be?

2. Your company has a Cloud Bigtable database that requires strong consistency, but it also requires high availability. You have implemented Cloud Bigtable replication and specified single-cluster routing in the app profile for the database. Some users have noted that they occasionally receive query results inconsistent with what they should have received. The problem seems to correct itself within a minute. What could be the cause of this problem?

3. You have been brought into a large enterprise to help with a data warehousing initiative. The first project of the initiative is to build a repository for all customer-related data, including sales, finance, inventory, and logistics. It has not yet been determined how the data will be used. What Google Cloud storage system would you recommend that the enterprise use to store that data?

4. You are using a UUID as the primary key in a Cloud Spanner database. You have noticed hotspotting that you did not anticipate. What could be the cause?

5. You have been tasked with migrating a MongoDB database to Cloud Spanner. MongoDB is a document database, similar to Cloud Firestore. You would like to maintain some of the document organization of the MongoDB design. What data type, available in Cloud Spanner, would you use to define a column that can hold a document-like structure?

6. A Cloud Spanner database is using a natural key as the primary key for a large table. The natural key is the preferred key by users because the values are easy to relate to other data. Database administrators notice that these keys are causing hotspots on Cloud Spanner nodes and are adversely affecting performance. What would you recommend in order to improve performance?

7. You are working for a financial services firm on a Cloud Bigtable database. The database stores equity and bond trading information from approximately 950 customers. Over 10,000 equities and bonds are tracked in the database. New data is received at a rate of 5,000 data points per minute. What general design pattern would you recommend?

8. Data is streaming into a BigQuery table. As the data arrives, it is added to a partition that was automatically created that day. Data that arrives the next day will be written to a different partition. The data modeler did not specify a column to use as a partition key. What kind of partition is being used?

9. What data formats are supported for external tables in Cloud Storage and Google Drive?

10. An. IoT startup has hired you to review their Cloud Bigtable design. The database stores data generated by over 100,000 sensors that send data every 60 seconds. Each row contains all the data for one sensor sent during an hour. Hours always start at the top of the hour. The row-key is the sensor ID concatenated to the hour of the day followed by the date. What change, if any, would you recommend to this design?

11. An application using a Cloud Spanner database has several queries that are taking longer to execute than the users would like. You review the queries and notice that they all involve joining three or more tables that are all related hierarchically. What feature of Cloud Spanner would you try in order to improve the query performance?

12. You are investigating long latencies in Cloud Bigtable query response times. Most queries finish in less than 20 ms, but the 99th percentile queries can take up to 400 ms. You examine a Key Visualizer heatmap and see two areas with bright colors indicating hotspots. What could be causing those hotspots?