A startup is launching a new mobile app backed by DynamoDB. Traffic patterns are completely unknown and could range from near-zero to a sudden viral spike. Which DynamoDB capacity mode best fits this scenario?
A) Provisioned capacity with Auto Scaling configured to a wide range.
B) On-Demand capacity mode.
C) Provisioned capacity with a fixed, high read/write capacity to handle any spike.
D) Reserved capacity purchased for a 1-year term.
Correct Answer: B
Explanation: On-Demand capacity mode automatically scales to handle workloads instantly with no capacity planning required, billing per request — ideal for new applications with unknown or highly unpredictable traffic patterns.
Why the others are wrong: (A) Auto Scaling on provisioned capacity is not instantaneous and can still throttle during very sudden spikes. (C) wastes money during low-traffic periods and still risks under-provisioning for an unknown spike size. (D) Reserved capacity requires committing to known, steady usage — the opposite of this scenario.