Examkin

Soru

Zorluk: Çok zorHigh-Performing Database Solutions

An organization is migrating its high-frequency financial ledger application to AWS. The system must meet two performance criteria: it must retrieve current transaction states with microsecond latency for active sessions, and it must write audit logs at a rate exceeding 80,00080,000 writes per second. The audit logs are queried chronologically by day. Which combination of database design and architectural strategies will meet these requirements? (Select TWO.)

  1. Store transaction states in Amazon DynamoDB and deploy Amazon DynamoDB Accelerator (DAX) to cache reads.Cevap
  2. For the audit log table, use a composite primary key consisting of a transaction UUID partition key and a timestamp sort key.Cevap
  3. C
    Configure the audit log DynamoDB table with a timestamp partition key to align with chronological query patterns.
  4. D
    Store transaction states in an Amazon RDS Multi-AZ DB cluster, using RDS Read Replicas as the primary target for application write operations during peak load periods.
  5. E
    Set the audit log DynamoDB table to Provisioned Capacity Mode with static write capacity units scaled to peak workload values to manage unpredictable, spiky traffic.

Cevap

The optimal strategy is to cache transaction state reads with DynamoDB Accelerator (DAX) to achieve microsecond latency, and to use a high-cardinality partition key (such as a transaction UUID) with a timestamp sort key for the audit logs to distribute write traffic evenly across DynamoDB partitions.
Caching transaction states using Amazon DynamoDB Accelerator (DAX) provides the required microsecond latency for active session lookups. Additionally, using a transaction UUID as the partition key and a timestamp as the sort key for the audit log table ensures write operations are evenly distributed across multiple physical partitions, preventing write bottlenecks while still supporting chronological queries.

Adım Adım Çözüm

1
Analyze read latency requirements.
Microsecond read latency requires an in-memory caching layer. Amazon DynamoDB Accelerator (DAX) provides microsecond latency for DynamoDB reads.
Standard DynamoDB queries operate in single-digit millisecond latency, which does not meet the microsecond requirement.
2
Analyze write scale and partition key design for the audit logs.
A rate of 80,00080,000 writes per second requires wide partition distribution to avoid hot keys.
DynamoDB partitions data based on the partition key. A sequential key (like a timestamp) routes all writes to a single partition, causing throttling.
3
Select partition and sort key strategies for chronological queries.
Use a transaction UUID as the partition key for write distribution, and the timestamp as the sort key for range-based chronological queries.
This allows query filters to run range operations on the sort key while maintaining uniform write distribution across partitions.

Anahtar Kavram

Designing high-performing AWS database architectures requires combining in-memory caching (such as DAX) for microsecond reads with proper partition key design (avoiding monotonically increasing attributes) to scale write throughput uniformly.
Bu soruyu puanla

Konular

High-Performing Database Solutions

Bu konudan daha fazla soru