Exam AWS Certified SysOps Administrator - Associate topic 1 question 244 discussion

If VolumeReadOps metric is low or dropping to less than 10% of its peak value during slowness, the issue isn't with disk I/O or IOPS. This suggests the CPU is being bottle necked. If the DB was slow due to disk I/O, then the I/O would be reading high (100%) not low. Since they mentioned this is a "burstable" "T" instance may be running in "Standard" mode, the easiest thing would be to enable "Unlimited" mode. In the real world you'd be looking at CPU utilization metric before you did that, but they trying to trick you with a dumb question.

VolumeReadOps metric is the metric of EBS. It drops 10% means CPU doesnt generate enough read traffic to EBS volume -> need to increase CPU rate -> only D is the possible answer.

A only

IOPS can only be solved via. EBS upgrade. Unlimited Mode does not help improve EBS IOPS.

Tenjin stated unlimited mode, but I do say that is correct because the question states "periodically", which means very so often. If it said "all he time" or "constantly" then you change to GP3.

Switching to a gp3 volume ensures consistent performance because: gp3 allows provisioning of baseline IOPS and throughput, independent of volume size. gp3 volumes offer lower cost with more predictable performance compared to gp2. You can increase IOPS and throughput for gp3 as needed to support the database workload without worrying about burst limits.

The unlimited mode does not resolve the VolumeReadOps issue.

The gp3 volumes provide a baseline performance of 3,000 IOPS and 125 MiB/s, regardless of the volume size, which can be increased if higher performance is needed1. This would address the issue of the VolumeReadOps metric dropping during periods of slow response, as gp3 volumes offer more consistent I/O performance than gp2 volumes12. Additionally, converting to gp3 may also be cost-effective, as it allows for independent provisioning of storage and IOPS3.

VolumeReadOps drop when burst credit are exhausted

https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/burstable-performance-instances-unlimited-mode-concepts.html#unlimited-mode-enabling

If it would be disk burst IOPS, the Baseline performance for this volume is 1002 IOPS which is much more then 10%. More like 30% gp2 volumes smaller than 1 TiB (and that are provisioned with less than 3,000 IOPS) can burst to 3,000 IOPS when needed for an extended period of time. 334 (min size for max throughput) Baseline IOPS: 1,002

gp2 has burst credits. If your workload continues to drive I/O traffic beyond the baseline performance, then your burst credit runs low. If your burst credit reaches zero, then these volume types get throttled at their baseline IOPS or throughput. https://repost.aws/knowledge-center/ebs-volume-burst-balance-low https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/general-purpose.html#gp2-performance

Option A (Convert the gp2 volume to a General Purpose SSD (gp3) EBS volume) is the correct option. The gp2 EBS volume type is designed for general-purpose workloads and provides a baseline performance and burst credits. The burst credits allow the volume to burst above the baseline performance when needed but are limited in capacity. If the volume depletes its burst credits, the performance can decrease, resulting in slow response times. On the other hand, gp3 is the next-generation general-purpose SSD EBS volume type that provides higher baseline performance, higher burst performance, and a new feature called "provisioned IOPS" that allows you to provision a consistent level of IOPS (Input/Output Operations Per Second) independent of volume size. By converting the gp2 volume to a gp3 volume, you will get higher baseline performance, which can help ensure consistently high performance for the database during periods of increased workload without relying solely on burst credits.

A is corect Option D (Activate unlimited mode on the EC2 instance) is related to burst performance on T2 and T3 instance types and does not directly address the I/O performance of the EBS volume. It is not the appropriate solution for addressing the slow response time issue. Based on the information provided in the question, option A is the best solution to ensure consistently high performance by upgrading the EBS volume to gp3. This change will provide higher baseline performance and better IOPS performance for the workload.

D is correct!!!!!!!

A is correct

Activate unlimited mode on the EC2 instance. A burstable performance instance configured as unlimited can sustain high CPU utilization for any period of time whenever required. The hourly instance price automatically covers all CPU usage spikes if the average CPU utilization of the instance is at or below the baseline over a rolling 24-hour period or the instance lifetime, whichever is shorter. https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/burstable-performance-instances-unlimited-mode.html