How should you model your lakehouse architecture to efficiently handle a high-volume, write-heavy IoT workload with millions of devices reporting every minute, while ensuring the ability to query data in near real-time?

Real Exam

Store raw data in a NoSQL database for write efficiency, periodically ETLing processed data into the lakehouse for analytical queries.

12.1%

Implement a sharded approach, creating separate tables for subsets of devices, and use a metastore to track shards for querying.

8.8%

Utilize a single wide table to store all IoT data, partitioned by device ID, leveraging Delta Lake‘s optimization for concurrent writes.

50.5%

Apply a micro-batching technique that combines streaming ingestion with periodic optimization (compaction and indexing) of the stored data for analysis.

28.6%

Databricks Certified Data Engineer - Professional

Get started today

Comments

How should you model your lakehouse architecture to efficiently handle a high-volume, write-heavy IoT workload with millions of devices reporting every minute, while ensuring the ability to query data in near real-time?