Microsoft Certified: Azure Solutions Architect ExpertDesign data storage solutionsMedium
A global manufacturing company needs to store sensor data from millions of IoT devices deployed worldwide. Each device sends small telemetry messages (JSON format) every few seconds. The data needs to be stored for up to 5 years for compliance and historical analysis. Critical requirements include extreme cost-effectiveness for storage, high write throughput, and the ability to perform simple queries (e.g., retrieve all data for a specific device within a time range) without complex joins. Which Azure data storage solution is the most appropriate?
- AAzure SQL Database
- BAzure Data Explorer (Kusto)
- CAzure Table Storage
- DAzure Cosmos DB (Core SQL API)
Show answer & explanationAnswer & explanation
Correct answer: C. Azure Table Storage
Azure Table Storage is a NoSQL key-value store optimized for storing massive amounts of structured, non-relational data at a low cost. It provides high write throughput and supports simple queries on primary key and partition key, making it ideal for IoT telemetry where complex joins are not required.
Why the other options are wrong
- A. Azure SQL Database is a relational database with higher costs and complexity for storing millions of small, simple IoT messages at scale.
- B. Azure Data Explorer is optimized for high-volume time-series data analytics and complex queries, but for simple key-value lookups and extreme cost-effectiveness for raw storage, Table Storage is more suitable.
- D. Azure Cosmos DB offers global distribution and low latency but is generally more expensive than Table Storage for simple key-value data with high volume and long retention.
Azure Table Storage
A NoSQL key-value store that stores large amounts of structured, non-relational data, providing high write throughput and cost-effectiveness for massive scale.
- Optimized for storing simple, schema-less data.
- Part of Azure Storage account, highly scalable.
- Cost-effective for high-volume data with simple query needs.
Memory trick: Tables are cheap for simple IoT, massive scale, no complex joins.