Learn
Practice
Projects
Free datasets
Resources
Toolkit
Pricing
Log in
Start free
Loading…
ETL / ELT Quiz (200 Questions)
← All topics
ETL / ELT Quiz (200 Questions)
200 exercises
·
10 free
01
1. In ETL, when does the Transformation step occur?
Free
→
02
2. ELT stands for:
Free
→
03
3. A key difference between ETL and ELT is:
Free
→
04
4. ETL is often chosen when:
Free
→
05
5. ELT is often preferred with modern cloud warehouses because:
Free
→
06
6. The main steps in ETL are:
Free
→
07
7. One advantage of ELT over ETL is:
Free
→
08
8. When choosing ETL vs. ELT, one key consideration is:
Free
→
09
9. ETL traditionally was favored in legacy systems because:
Free
→
10
10. One drawback of ETL is:
Free
→
11
11. ELT leverages the target system's resources to:
Locked
→
12
12. Both ETL and ELT ultimately aim to:
Locked
→
13
13. ETL pipelines often run on a schedule to:
Locked
→
14
14. ELT is more flexible when requirements change because:
Locked
→
15
15. ELT often aligns well with modern data lake or warehouse strategies because:
Locked
→
16
16. One disadvantage of ELT is:
Locked
→
17
17. Deciding between ETL and ELT depends on:
Locked
→
18
18. ELT often pairs well with:
Locked
→
19
19. ETL pipelines might struggle with changing requirements because:
Locked
→
20
20. Ultimately, ETL and ELT are chosen based on:
Locked
→
21
21. Data extraction involves:
Locked
→
22
22. Incremental extraction means:
Locked
→
23
23. Handling schema changes in extraction involves:
Locked
→
24
24. Extracting from relational databases often uses:
Locked
→
25
25. APIs for extraction may require handling:
Locked
→
26
26. Files (CSV, JSON) extraction involves:
Locked
→
27
27. Handling authentication during extraction might involve:
Locked
→
28
28. When a source system is slow, extraction strategies might include:
Locked
→
29
29. Ensuring data integrity during extraction might involve:
Locked
→
30
30. Source system locks and concurrency issues during extraction can be mitigated by:
Locked
→
31
31. Metadata extraction during extraction phase involves:
Locked
→
32
32. Logging extraction events helps by:
Locked
→
33
33. Scheduling extractions often involves:
Locked
→
34
34. Dealing with unreliable sources may involve:
Locked
→
35
35. Handling different file formats (CSV, JSON, XML) requires:
Locked
→
36
36. Some ETL tools come with built-in source connectors to:
Locked
→
37
37. Minimizing network transfers during extraction might involve:
Locked
→
38
38. Validating extracted data ensures:
Locked
→
39
39. Metadata like source timestamps help by:
Locked
→
40
40. Ensuring a stable extraction process might mean:
Locked
→
41
41. Data transformation often includes:
Locked
→
42
42. Converting date formats and normalizing data units is part of:
Locked
→
43
43. Business rules in transformations might include:
Locked
→
44
44. SQL-based transformations in ELT approach use:
Locked
→
45
45. Handling Slowly Changing Dimensions (SCD) in transformations means:
Locked
→
46
46. Joining multiple data sources in transformation steps is common to:
Locked
→
47
47. Applying data quality checks mid-transformation ensures:
Locked
→
48
48. Debugging transformation logic often involves:
Locked
→
49
49. Using frameworks like Spark for transformation helps with:
Locked
→
50
50. Ensuring transformations are idempotent means:
Locked
→
51
51. Applying aggregations (e.g., sum, avg) during transformations helps to:
Locked
→
52
52. Handling character encoding issues (UTF-8 vs. ASCII) in transformations ensures:
Locked
→
53
53. Versioning transformation logic means:
Locked
→
54
54. Pushdown transformations refer to:
Locked
→
55
55. Idempotent transformations mean if rerun:
Locked
→
56
56. Reusability in transformations can be achieved by:
Locked
→
57
57. Handling schema evolution during transformation involves:
Locked
→
58
58. Transformation performance optimization may include:
Locked
→
59
59. Debugging transformation errors might use:
Locked
→
60
60. Once transformations are finalized:
Locked
→
61
61. Data loading involves:
Locked
→
62
62. Bulk loading can improve performance by:
Locked
→
63
63. Incremental loading (upserts) means:
Locked
→
64
64. Managing indexes during load may involve:
Locked
→
65
65. Timing loads off-peak hours can:
Locked
→
66
66. Transactions during load ensure:
Locked
→
67
67. Verifying load success might involve:
Locked
→
68
68. Partitioned loading improves performance by:
Locked
→
69
69. Using target-specific load utilities (e.g., COPY command in Redshift) can:
Locked
→
70
70. Handling load failures might involve:
Locked
→
71
71. Distinguishing between full refresh and incremental refresh loading means:
Locked
→
72
72. Post-load validations ensure:
Locked
→
73
73. Notifications on load completion or errors help by:
Locked
→
74
74. Balancing load tasks means:
Locked
→
75
75. Late-arriving data might be handled by:
Locked
→
76
76. Atomicity of loads ensures:
Locked
→
77
77. Archiving or purging old data during load cycles is done to:
Locked
→
78
78. Ensuring consistency and atomicity of loads might require:
Locked
→
79
79. In ELT scenarios, loading raw data first allows:
Locked
→
80
80. After successful load:
Locked
→
81
81. Traditional ETL tools (Informatica, Talend) often:
Locked
→
82
82. Modern ELT tools integrate with cloud warehouses like:
Locked
→
83
83. Orchestration tools (Airflow, Luigi) help by:
Locked
→
84
84. SaaS integration platforms (Fivetran, Stitch) often:
Locked
→
85
85. Using Python scripts for ETL can be advantageous for:
Locked
→
86
86. Dockerizing ETL jobs provides:
Locked
→
87
87. CI/CD pipelines for ETL code mean:
Locked
→
88
88. On-prem ETL tools vs. cloud-native solutions differ in:
Locked
→
89
89. Data virtualization tools help by:
Locked
→
90
90. Evaluating open-source vs. commercial ETL solutions involves:
Locked
→
91
91. Performance benchmarks among ETL tools help to:
Locked
→
92
92. Leveraging Spark or Flink in transformations is useful for:
Locked
→
93
93. DataOps integrates ETL/ELT with:
Locked
→
94
94. Integration with data catalogs helps ETL/ELT by:
Locked
→
95
95. Using message queues (Kafka) in extraction helps by:
Locked
→
96
96. ETL in a microservices architecture might mean:
Locked
→
97
97. Impact of orchestration tool’s scheduling features:
Locked
→
98
98. Version control of ETL scripts and configs helps with:
Locked
→
99
99. Selecting a tool based on data volume and complexity means:
Locked
→
100
100. Considering team skill sets for tool selection means:
Locked
→
101
101. Identifying bottlenecks in ETL/ELT pipelines often involves checking:
Locked
→
102
102. Parallelization strategies might include:
Locked
→
103
103. Partitioning data for parallel processing helps by:
Locked
→
104
104. Efficient file formats like Parquet or ORC improve performance by:
Locked
→
105
105. Compressing data before transfer reduces:
Locked
→
106
106. Memory management for large ETL jobs can be improved by:
Locked
→
107
107. Caching intermediate results might help if:
Locked
→
108
108. Query optimization in ELT scenarios includes:
Locked
→
109
109. Minimizing unnecessary data movement means:
Locked
→
110
110. Scheduling ETL jobs off-peak can improve performance by:
Locked
→
111
111. Monitoring runtime metrics (CPU, memory, throughput) helps:
Locked
→
112
112. Choosing incremental over full loads can improve performance by:
Locked
→
113
113. Using columnar storage in the target system helps because:
Locked
→
114
114. Monitoring runtime with trend analysis helps:
Locked
→
115
115. Retry and backoff strategies during extraction and loading help by:
Locked
→
116
116. Eliminating unnecessary transformations means:
Locked
→
117
117. Code profiling in ETL scripts helps by:
Locked
→
118
118. Reducing unnecessary data movement (ELT vs. ETL) can improve performance by:
Locked
→
119
119. Adopting streaming ETL for continuous processing improves performance for:
Locked
→
120
120. Automating performance regression tests means:
Locked
→
121
121. Implementing validation checks at extraction ensures:
Locked
→
122
122. Data cleansing involves:
Locked
→
123
123. Standardizing reference data during transformation helps:
Locked
→
124
124. Data lineage means:
Locked
→
125
125. Auditing changes and maintaining historical versions of data allows:
Locked
→
126
126. Data quality metrics (completeness, consistency) help by:
Locked
→
127
127. Error handling pipelines (quarantine bad records) means:
Locked
→
128
128. Role-based access controls in ETL/ELT governance ensure:
Locked
→
129
129. Metadata management helps by:
Locked
→
130
130. Ensuring consistency between source and target schemas prevents:
Locked
→
131
131. Self-service data quality checks mean:
Locked
→
132
132. SLA definitions for data timeliness and correctness ensure:
Locked
→
133
133. Logging quality metrics allows:
Locked
→
134
134. Governance frameworks (like DAMA) applied to ETL/ELT mean:
Locked
→
135
135. Aligning ETL/ELT practices with organizational policies ensures:
Locked
→
136
136. Continuous improvement cycles for data quality involve:
Locked
→
137
137. Ensuring completeness means:
Locked
→
138
138. Consistency checks ensure:
Locked
→
139
139. Accuracy checks might compare extracted data against:
Locked
→
140
140. Integrating with data governance tools means:
Locked
→
141
141. Encrypting data in transit ensures:
Locked
→
142
142. Using secure protocols (TLS/SSL) for data extraction from APIs prevents:
Locked
→
143
143. Masking or tokenizing PII fields in transformations ensures:
Locked
→
144
144. Applying column-level encryption or hashing can:
Locked
→
145
145. Strict access controls on ETL pipelines means:
Locked
→
146
146. Complying with regulations like GDPR may involve:
Locked
→
147
147. Auditing and logging who accessed ETL data ensures:
Locked
→
148
148. Using secrets managers for credentials instead of hardcoding prevents:
Locked
→
149
149. Minimizing data movement of sensitive records may mean:
Locked
→
150
150. Regular compliance audits and ETL process reviews ensure:
Locked
→
151
151. Using VPCs or private networking for data transfers ensures:
Locked
→
152
152. Complying with HIPAA for healthcare data might require:
Locked
→
153
153. Separation of duties in ETL/ELT management means:
Locked
→
154
154. Using temporary credentials or tokens instead of static keys enhances security by:
Locked
→
155
155. Applying principle of least privilege to ETL system accounts means:
Locked
→
156
156. Sanitizing logs and debug info means:
Locked
→
157
157. Ensuring data disposal and retention policies are followed means:
Locked
→
158
158. Using anonymized or synthetic test data prevents:
Locked
→
159
159. Regular compliance audits might involve:
Locked
→
160
160. Storing sensitive data only in encrypted form reduces risk if:
Locked
→
161
161. Real-time ETL differs from batch ETL by:
Locked
→
162
162. Change Data Capture (CDC) techniques detect:
Locked
→
163
163. Tools like Debezium or Attunity assist with:
Locked
→
164
164. Using Kafka or Kinesis in streaming ETL means:
Locked
→
165
165. Micro-batching vs. true streaming differs by:
Locked
→
166
166. Handling out-of-order events in streaming ETL often requires:
Locked
→
167
167. Ensuring exactly-once delivery in streaming ETL means:
Locked
→
168
168. State management in streaming transformations involves:
Locked
→
169
169. Windowing functions (tumbling, sliding windows) in streaming help by:
Locked
→
170
170. Dealing with backpressure and rate limiting ensures:
Locked
→
171
171. Transforming on the fly vs. storing raw events first is a choice between:
Locked
→
172
172. Choosing storage sinks for streaming outputs (NoSQL, data lakes) depends on:
Locked
→
173
173. Monitoring streaming pipelines for lag and throughput helps by:
Locked
→
174
174. Recovery and fault tolerance in streaming ETL might use:
Locked
→
175
175. Schema evolution in a streaming environment means:
Locked
→
176
176. Watermarks in event-time processing help by:
Locked
→
177
177. Integration with stream processing frameworks (Flink, Spark Structured Streaming) means:
Locked
→
178
178. Real-time alerting on data quality in streams allows:
Locked
→
179
179. Balancing latency vs. completeness means:
Locked
→
180
180. Continuous integration and deployment for streaming pipelines means:
Locked
→
181
181. Unit tests for individual transformation logic mean:
Locked
→
182
182. Integration tests across the entire ETL pipeline ensure:
Locked
→
183
183. Using mock sources and targets in tests allows:
Locked
→
184
184. Regression testing ensures:
Locked
→
185
185. Load testing ETL pipelines helps by:
Locked
→
186
186. Synthetic test data generation helps test:
Locked
→
187
187. Integrating tests with CI/CD means:
Locked
→
188
188. Automated error detection and retries in pipelines means:
Locked
→
189
189. Monitoring tools (e.g., Airflow UI) provide:
Locked
→
190
190. Alerting on job failures or SLA breaches means:
Locked
→
191
191. Logging best practices (structured logs) aid in:
Locked
→
192
192. Version control of ETL pipeline definitions means:
Locked
→
193
193. Auditing job runs (who triggered them, when) provides:
Locked
→
194
194. Scalability tests ensure:
Locked
→
195
195. Using checksums and row counts for validation in tests means:
Locked
→
196
196. Canary deployments for ETL changes mean:
Locked
→
197
197. Performance metrics (CPU, memory) in monitoring help identify:
Locked
→
198
198. Trend analysis in monitoring dashboards means:
Locked
→
199
199. Security scans on ETL code (linters, secret detection) ensure:
Locked
→
200
200. Continuous improvement loops from monitoring data means:
Locked
→
