An application needs DynamoDB reads that always reflect the latest committed write, even if a write just occurred milliseconds ago. Which read consistency type should be used?
- AEventually Consistent Reads — might return stale data for up to a second after a write.
- BStrongly Consistent Reads — always returns the most up-to-date data, reflecting all successful writes.✓ Correct
- CTransactional Reads — for multi-item atomic operations, not a consistency level for regular reads.
- DBatch Reads — for reading multiple items efficiently, not a consistency model.
DynamoDB offers two read consistency options: Eventually Consistent (default, 50% the cost of strongly consistent) may return slightly stale data. Strongly Consistent reads always reflect the latest data but consume more read capacity units and have slightly higher latency. Strong consistency is not available for Global Secondary Indexes.
A company's DynamoDB table experiences unpredictable traffic with sudden spikes several times a day. The team wants AWS to automatically manage read and write capacity. Which capacity mode is BEST?
- AOn-Demand Capacity Mode — automatically scales to accommodate traffic with no capacity planning required.✓ Correct
- BProvisioned Capacity with manual settings — requires predicting peak traffic and setting fixed RCUs/WCUs.
- CProvisioned Capacity with Auto Scaling — reactive, may not scale fast enough for sudden spikes.
- DDynamoDB Reserved Capacity — for cost savings on provisioned capacity, not automatic scaling.
On-Demand mode charges per request and instantly accommodates any traffic level without throttling. It is ideal for unpredictable workloads. Provisioned mode with Auto Scaling reacts to sustained traffic changes but may throttle during sudden spikes. On-Demand can be up to 2.5× more expensive per request than provisioned at steady-state loads.
A gaming leaderboard application needs to cache frequently accessed DynamoDB items with single-digit microsecond latency. Which AWS service is purpose-built for this use case?
- AAmazon ElastiCache Redis — a general-purpose cache but requires custom integration with DynamoDB.
- BAmazon DynamoDB Accelerator (DAX) — in-memory cache fully compatible with DynamoDB API, microsecond latency.✓ Correct
- CAmazon CloudFront — a CDN for caching HTTP responses, not DynamoDB items.
- DAmazon MemoryDB for Redis — a Redis-compatible database, not a DynamoDB cache.
DAX is a fully managed, highly available, in-memory cache for DynamoDB. It is API-compatible with DynamoDB, so minimal code changes are needed. DAX delivers up to 10× read performance improvement and reduces DynamoDB read capacity consumption. Ideal for read-heavy workloads like leaderboards, product catalogs, and session stores.
Which DynamoDB feature enables active-active multi-region replication, allowing reads and writes in multiple AWS regions simultaneously?
- ADynamoDB Streams — captures a time-ordered sequence of item-level changes in a table.
- BDynamoDB Global Tables — multi-region, multi-active replication with automatic conflict resolution.✓ Correct
- CDynamoDB Cross-Region Replication (manual Lambda-based) — an older workaround, not the intended solution.
- DDynamoDB Multi-AZ — DynamoDB is already replicated across 3 AZs by default within a region.
DynamoDB Global Tables create replica tables in multiple AWS regions. All replicas are writable (active-active). Changes are propagated with sub-second replication latency. Conflict resolution uses "last writer wins" semantics. Global Tables require DynamoDB Streams to be enabled and are built on top of it internally.
A financial application needs to atomically debit one DynamoDB item and credit another item in a different table. If either operation fails, neither should be committed. Which DynamoDB feature provides this?
- ABatchWriteItem — writes up to 25 items but does NOT guarantee atomicity; partial failures are possible.
- BDynamoDB Transactions (TransactWriteItems) — provides all-or-nothing ACID operations across multiple items and tables.✓ Correct
- CDynamoDB Streams — captures change events but does not provide transactional writes.
- DConditional Writes — apply conditions to single-item writes but cannot span multiple items atomically.
DynamoDB Transactions support ACID properties. TransactWriteItems performs up to 100 write actions across one or more tables atomically. If any condition fails, the entire transaction is rolled back. Transactions consume 2× the normal write capacity units. They are ideal for financial operations, inventory management, and order processing.