CDC and Real-Time — Learn free

Self-check

Questions 1–10 of 10

1. What does CDC stand for in data processing?
Question 1A. Central Data CatalogueB. Change Data CaptureC. Continuous Data CopyD. Controlled Data CacheE. Consolidated Data Centre
2. What does a CDC process pick up and push to the next layer?
Question 2A. The whole table every timeB. Only the table metadataC. A random sample of rowsD. Only the data that has changedE. Only rows older than a day
3. Which of the following is NOT one of the kinds of change that CDC captures?
Question 3A. A table being re-indexedB. A new row being addedC. An existing row being updatedD. A previously added row being deletedE. None of these; all are captured
4. Why is CDC mostly not run in Real-Time even though it could be?
Question 4A. Because databases forbid real-time readsB. Because CDC only works in overnight batchesC. Because Real-Time plus CDC needs a lot of memory processing, which carries a heavy TCO, so it depends on the ROI of the business caseD. Because transaction logs are only written dailyE. Because real-time data has no previous state
5. Which CDC method is described as the slowest and requiring the heaviest processing?
Question 5A. Reading the transaction logB. Using date columns in source tablesC. Capturing changes at the application levelD. Streaming from CCTVE. Keeping the previous dataset in one table, loading the new dataset in another and comparing them
6. Which CDC method is recommended and most commonly used?
Question 6A. Comparing two full copies of the tableB. Using date columns in every source tableC. Reading the transaction log generated by insert, update and delete operationsD. Capturing changes at the application levelE. Polling the source tables every second
7. What is the drawback of the CDC method that relies on date columns in source tables?
Question 7A. It requires the source data model to have date fields in each table, which may need extensive development effort, and it has a processing impact on the sourceB. It cannot detect deleted rowsC. It only works with transaction logsD. It is the most recent method and few solutions support itE. It requires an in-memory database
8. Which statement describes the application-level CDC method?
Question 8A. It compares two snapshots of each tableB. It is the oldest and most widely adopted methodC. It relies on date columns in the sourceD. It has a heavy impact on source I/OE. It is the most recent method, used by very few solutions, in which the application pushes each write for CDC processing with no impact on the database or I/O
9. Why does the pipeline carrying a Facebook LIKE downstream have nothing to do with CDC?
Question 9A. Because Facebook does not use databasesB. Because it always appends from the application level rather than capturing a change to a previous stateC. Because LIKE data is processed in batchesD. Because CDC only applies to deletionsE. Because CDC is only used in data warehouses
10. What must organisations planning for Real-Time processing evaluate?
Question 10A. Which BI tool to buyB. How many batches to run per dayC. Whether to switch to a NoSQL databaseD. Whether CDC is required at allE. Whether to disable transaction logs