You receive per‑second discrete events (sensor ID + small discrete fields) from 1000 motion sensors across offices; analysts will join this with owner and location data. Which database type should you use?

NoSQL databases are ideal for this scenario because they handle high volumes of discrete, schema-less data (per-second events from 1000 sensors) and offer excellent scalability for ingestion and retrieval. The ability to easily join this data with other owner and location information, which might also be semi-structured, further supports a NoSQL approach. Relational databases would struggle with the high write velocity and the potentially flexible schema of sensor data, requiring complex schema management and potentially impacting performance. Flat files lack the querying capabilities and scalability needed for real-time analysis. Blobstore is for unstructured large objects, not discrete, queryable events.