Question

Difficulty: MediumAzure Cosmos DB and NoSQL Solutions

A logistics company is designing a telemetry ingestion backend for 100,000100,000 active delivery vehicles. Each vehicle uploads coordinate and sensor data every 1010 seconds. The storage solution must be designed to meet the following requirements:
- Regional disaster resilience must be guaranteed with a recovery point objective (RPORPO) of zero.
- Write-path hot partitioning must be prevented under high-throughput ingestion.
- Security credentials for read access must be easily revocable.

Which configuration should you recommend?

  1. A
    Configure an Azure Cosmos DB account with a single write region and a secondary replica region using Strong consistency, and partition the container using TelemetryDate.
  2. Configure an Azure Cosmos DB account with a single write region and a secondary replica region using Strong consistency, partition the container using VehicleID, and utilize a token broker to issue short-lived Cosmos DB resource tokens for client access.Answer
  3. C
    Configure an Azure Storage Table with Locally Redundant Storage (LRS) replication, partition by VehicleID, and use a token broker to issue short-lived resource tokens.
  4. D
    Configure an Azure Cosmos DB account with a single write region and a secondary replica region using Strong consistency, partition the container using VehicleID, and authorize clients using ad-hoc Shared Access Signature (SAS) tokens configured with a five-year validity period without a stored access policy.

Answer

Configure an Azure Cosmos DB account with a single write region and a secondary replica region using Strong consistency, partition the container using VehicleID, and utilize a token broker to issue short-lived Cosmos DB resource tokens for client access.
The configuration utilizing a single write region with a secondary replica under Strong consistency satisfies the RPO = 0 requirement by executing synchronous replication. Using the high-cardinality VehicleID as the partition key ensures that writes are evenly distributed, avoiding hot partitions. Generating short-lived resource tokens via a token broker ensures that read access can be easily managed and revoked.

Step-by-Step Solution

1
Select the database service and replication configuration to meet the disaster recovery constraint.
Azure Cosmos DB configured with a single write region and a secondary replica using Strong consistency.
Strong consistency guarantees that write operations are committed synchronously to the secondary replica before completing, ensuring a recovery point objective (RPO) of zero.
2
Select a partition key that distributes the write workload evenly.
Partition key: VehicleID.
With 100,000100,000 active vehicles, VehicleID has high cardinality. This distributes the ingestion throughput across many logical partitions, preventing write-path hot partitions.
3
Determine the access control mechanism for client applications.
Utilize a token broker to issue short-lived Cosmos DB resource tokens.
Resource tokens allow fine-grained access control to specific Cosmos DB resources and expire quickly, making them easily revocable and secure.

Key Concept

Designing partition keys and consistency configurations in Azure Cosmos DB to meet scalability, disaster recovery, and security requirements.
Rate this question