DEA-C01 Practice Question 34: DynamoDB TTL Expiry
AWS Certified Data Engineer – Associate (DEA-C01) · Test 4 · question 4 of 10
An Amazon DynamoDB table stores device telemetry events. Only the last 30 days of events are ever read; older items are dead weight, and the table's storage bill grows every month. The team wants expired items removed automatically, without consuming provisioned write capacity and without running a maintenance job.
Which approach should they use?
Show answer and explanation
Answer:
TTL is described as "a cost-effective method for deleting items that are no longer relevant." You enable it on the table, nominate an attribute, and store a per-item expiration timestamp in it; "DynamoDB automatically deletes expired items within a few days of their expiration time, without consuming write throughput." That matches all three constraints: automatic, no write capacity consumed, and no job to operate. The timestamp must be a Number in Unix epoch seconds; items whose TTL attribute is another type are simply ignored by the TTL process.
Option B is the manual version and is expensive twice over. A nightly Scan reads the entire table and consumes read capacity, and every delete consumes write capacity, precisely what the requirement rules out, plus there is now a job to monitor.
Option C works but shifts complexity into the application and the operations process. Readers must know which table or tables to query for a given window, cross-month queries span two tables, and a human or automation has to remember to create and drop tables on schedule.
Option D misuses a backup feature and is destructive. A point-in-time restore creates a new table at a past state, which would discard the last 30 days of writes, the only data anyone actually reads, rather than removing old items.
Reference: Using time to live (TTL) in DynamoDB