Northern Trail Outfitters is streaming IoT data from connected devices to a cloud database. Every 24 hours. 100,000 records are generated. NIO employees will need to see these lol records within Salesforce and generate weekly reports on it. Developers may also need to write programmatic logic to aggregate the records and incorporate them into workflows. Which data pattern will allow a data architect to satisfy these requirements, while also keeping limits in mind?
A.
Bidirectional integration
B.
Unidirectional integration
C.
Virtualization
D.
Persistence
C.
Virtualization
Explanation:
Virtualization (via Salesforce Connect and External Objects) allows real-time access to IoT data stored externally without replicating it in Salesforce. This avoids storage limits (100k daily = 36.5M/year) and preserves API capacity. Bidirectional (A) or unidirectional (B) integration would require data replication, consuming excessive storage and ETL resources. Persistence (D) (storing data in Salesforce) would hit storage caps and slow reporting. With virtualization:
Users see live IoT data in Salesforce via OData adapters.
Weekly reports use cross-org queries (combining Salesforce and external data).
Developers access records programmatically via SOQL on external objects.
Workflows trigger based on real-time data changes.
This pattern minimizes Salesforce storage costs, adheres to governor limits, and meets all access requirements.