SAA-C03 Practice Question 98: EFS Elastic Throughput Spiky
AWS Certified Solutions Architect – Associate (SAA-C03) · Test 10 · question 8 of 10
A media team stores 200 GiB of working files on an Amazon EFS file system that uses Bursting throughput mode. Several times a day, a batch of jobs reads and writes at high speed for a few minutes, and during these bursts the jobs slow down because the file system's throughput is limited. Traffic between bursts is very low, and the team cannot predict how high each burst will be. They want to pay only for the throughput they actually use.
What should the architect do?
Show answer and explanation
Answer:
In Bursting mode, throughput scales with the amount of data stored, so a small file system has a low baseline and depends on burst credits that a heavy workload can use up. Elastic throughput mode automatically scales throughput up and down with the workload and charges for the throughput actually consumed. It is designed for spiky or unpredictable access patterns, which matches this file system, and it needs no capacity forecast.
Option A changes the wrong dimension. The performance mode governs the latency and IOPS profile, not the throughput limit, Max I/O cannot be combined with Elastic throughput, and a file system's performance mode cannot be changed after it is created.
Option B improves availability and lets clients mount from their own Availability Zone, but the file system's throughput limit is set by its throughput mode, not by the number of mount targets.
Option C would remove the slowdown, but Provisioned mode bills for the throughput configured at all times, so the team would pay for the peak around the clock while the file system is idle most of the day. It also requires a forecast the team says it cannot make.
Reference: Amazon EFS performance specifications