Examkin

Question

Difficulty: MediumAzure Cosmos DB and NoSQL Solutions

An enterprise is deploying a global fleet-tracking system for international cargo ships to monitor fuel efficiency and environmental metrics. The database must support a write-intensive workload with an 80:2080:20 write-to-read ratio, ingesting data globally across the East US, West Europe, and East Asia regions. The system requires sub-10ms latency for both reads and writes, a 99.999%99.999\% availability SLA for both read and write operations, and must survive regional datacenter outages without data loss. Additionally, database backup data must be resilient against regional disasters. Which two configurations should you recommend to meet these requirements?

  1. Configure the Azure Cosmos DB account to use multi-region writes across East US, West Europe, and East Asia.Answer
  2. Partition the Cosmos DB container using a high-cardinality composite key consisting of the ship identifier and the transmission hour.Answer
  3. C
    Partition the Cosmos DB container using the continent code of the current location as the partition key.
  4. D
    Configure the periodic backup storage redundancy of the Cosmos DB account to use Locally Redundant Storage (LRS).

Answer

To meet the requirements, you must configure the Azure Cosmos DB account to use multi-region writes and partition the container using a high-cardinality composite key of the ship identifier and transmission hour.
Enabling multi-region writes ensures writes are processed locally in East US, West Europe, and East Asia, keeping latency under 10ms and fulfilling the 99.999% SLA. A high-cardinality composite key of the ship identifier and transmission hour spreads data and Request Units (RUs) uniformly across physical partitions to accommodate the write-intensive telemetry stream.

Step-by-Step Solution

1
Evaluate the latency and availability SLA requirements.
To achieve sub-10ms write latency globally and a 99.999%99.999\% read/write availability SLA, the system must write to local regional endpoints rather than routing all writes to a single primary region.
Multi-region writes in Azure Cosmos DB satisfy the latency limit and elevate the write availability SLA to 99.999%99.999\%.
2
Determine the partitioning strategy to avoid hot partitions.
Select a high-cardinality partition key that distributes writes uniformly across partitions over time.
Using a composite key like ship identifier and transmission hour ensures write operations are evenly spread, avoiding the bottlenecks of low-cardinality keys.
3
Analyze backup redundancy needs under disaster recovery constraints.
Discard Locally Redundant Storage (LRS) for backups and ensure geo-redundant or zone-redundant storage is configured.
LRS keeps data in a single facility and cannot survive regional datacenter outages, failing the disaster resilience goal.

Key Concept

Azure Cosmos DB multi-region writes, high-cardinality partitioning, and backup redundancy options for high-availability globally distributed databases.
Rate this question

Topics

Azure Cosmos DB and NoSQL Solutions

More questions from this topic