Data Lakehouse — Learn free

Self-check

Questions 1–10 of 10

1. In simple terms, what is a data lakehouse?
Question 1A. A data lake without any governanceB. Data warehouse + data lake, combining the strengths of bothC. A data warehouse hosted in the cloudD. A NoSQL database for streamingE. A virtual layer over operational systems
2. Around when did the data lakehouse concept emerge?
Question 2A. Around 1990B. Around 2000C. Around 2010D. Around 2020E. Around 1975
3. What long-standing weakness of traditional data lakes does the lakehouse address?
Question 3A. They could not store unstructured dataB. They lacked strong transactional guarantees, often leading to inconsistent or unreliable data setsC. They were too expensive to scaleD. They could only run batch workloadsE. They required 3NF modelling
4. Which of the following is NOT one of the capabilities a data lakehouse offers?
Question 4A. ACID transactionsB. Upsert and delete operationsC. Schema-on-write supportD. Tight coupling of storage and computeE. Batch and streaming compatibility
5. What is the significance of open table formats such as Parquet in a data lakehouse?
Question 5A. They encrypt the data automaticallyB. They make data natively accessible to analytical and data science toolsC. They restrict access to a single vendor's toolsD. They convert data into relational tablesE. They remove the need for metadata
6. What architectural breakthrough of the lakehouse addresses the scalability bottlenecks of earlier systems?
Question 6A. Storing everything in memoryB. The decoupling of storage and compute, with compute clusters scaled independentlyC. Replacing SQL with a new query languageD. Moving all data back into OLTP databasesE. Eliminating metadata management
7. Which foundation does a data lakehouse use for storage?
Question 7A. Relational databases such as OracleB. Inexpensive, durable object storage such as Amazon S3, Azure Data Lake Storage or Google Cloud StorageC. Tape archivesD. Local disks on each compute nodeE. In-memory caches
8. Why is object storage alone not sufficient for a lakehouse, and what fills the gap?
Question 8A. It is too slow; the gap is filled by in-memory cachingB. It lacks reliability and governance; technologies such as Delta Lake add transactional integrity and metadata management on topC. It cannot hold unstructured data; the gap is filled by HDFSD. It is too expensive; the gap is filled by tapeE. It does not support Parquet; the gap is filled by CSV
9. How does schema-on-write support in a lakehouse compare with schema-on-read designs?
Question 9A. It allows weaker data contractsB. It allows stronger data contractsC. It removes the need for any schemaD. It only applies to unstructured dataE. It is identical to schema-on-read
10. Why is the lakehouse model gaining momentum even though many organisations still maintain Hadoop-based data lakes?
Question 10A. Because it is free of chargeB. Because it unifies analytical, machine learning and streaming workloads in a single architecture, reducing overlapping silos and simplifying governanceC. Because it eliminates the need for object storageD. Because it only supports structured dataE. Because regulators mandate it