Microsoft Certified: Fabric Analytics Engineer AssociateImplement and manage semantic models (30-35%)Medium
A data engineer is configuring a semantic model refresh in Microsoft Fabric. The underlying data source is a large transactional database that updates constantly throughout the day. The business requires the latest data to be available in reports with minimal latency, ideally within minutes of changes occurring in the source. Which refresh type is most suitable for this requirement?
- AScheduled Refresh
- BIncremental Refresh
- CReal-time Refresh (DirectQuery)
- DFull Refresh
Show answer & explanationAnswer & explanation
Correct answer: C. Real-time Refresh (DirectQuery)
Real-time refresh, typically achieved through DirectQuery or Direct Lake mode, ensures that reports always reflect the latest data directly from the source without needing a separate refresh process, which is ideal for minimal latency requirements.
Why the other options are wrong
- A. Scheduled refresh involves periodic updates, which will always introduce some delay between source changes and report availability.
- B. Incremental refresh only loads new or changed data, improving efficiency, but still involves a scheduled refresh cycle, introducing some latency.
- D. Full refresh reloads all data, which is time-consuming and resource-intensive for constantly updating, large datasets.
Real-time Refresh
A refresh strategy where data in a semantic model is queried directly from the source system at the time of report interaction, ensuring the freshest possible data.
- Achieved primarily through DirectQuery or Direct Lake storage modes.
- Minimizes data latency to near real-time.
- Performance depends heavily on the underlying data source's query capabilities.
Memory trick: For instant updates, query directly; for speed, import and schedule.