SOA-C02 Question 326
Select 2You are a SysOps Administrator managing an application that uses Amazon EBS volumes for data storage. The application is experiencing periodic performance bottlenecks during high traffic periods. Upon reviewing Amazon CloudWatch metrics, you notice that the VolumeQueueLength metric is consistently high during these times. Which of the following actions should you take to resolve the performance issue? (Choose TWO)
- A
Switch the EBS volume type to a Provisioned IOPS SSD (io2) volume.
- B
Increase the size of the EBS volume to improve throughput.
- C
Move the EBS volume to a different Availability Zone.
- D
Enable EBS Multi-Attach to use multiple instances for higher throughput.
- E
Distribute the workload across multiple EBS volumes.
Show answer and explanation
Correct answers: A, E
Explanation
A high VolumeQueueLength indicates that the EBS volume is unable to handle the I/O requests due to insufficient performance capacity. Switching to a Provisioned IOPS SSD (io2) volume ensures consistent high IOPS performance, while distributing the workload across multiple EBS volumes reduces the load on a single volume and improves overall performance efficiency. These are the most effective solutions in this scenario.
- A. Correct.
Switching to a Provisioned IOPS SSD (io2) volume is a valid solution as it provides consistent and high-performance IOPS, which can help address high VolumeQueueLength issues caused by insufficient IOPS capacity.
- B. Incorrect.
Increasing the size of the EBS volume does not directly address high VolumeQueueLength, as increasing size improves throughput only for Throughput Optimized HDD (st1) or Cold HDD (sc1) volumes, but not IOPS performance.
- C. Incorrect.
Moving the EBS volume to a different Availability Zone does not improve performance efficiency. This action is unrelated to addressing high VolumeQueueLength.
- D. Incorrect.
EBS Multi-Attach is used to attach a volume to multiple instances but does not increase performance or throughput for a single instance. It is not relevant for resolving high VolumeQueueLength issues.
- E. Correct.
Distributing the workload across multiple EBS volumes is an effective way to increase performance efficiency because it reduces the load on a single volume, thereby lowering VolumeQueueLength.