By using AWS re:Post, you agree to the AWS re:Post Terms of Use

AWS Elastic Beanstalk CPU utilization metric for burstable machine - does it make sense?

0

Hi, For an AWS Elastic Beanstalk environment using Enhanced Health Monitoring and a burstable machine type (t3a family), we were a bit surprised to find CPU utilization still counting regarding the environment health metric. We understand that even for a burstable machine type, permanent very high CPU utilization rates should not be considered normal, but shouldn't the Beanstalk environment health metric be somewhat more relaxed? Besides that, where can I find the exact environment health component metric's thresholds?

Thanks!

asked a month ago30 views
1 Answer
1

While AWS Elastic Beanstalk does use CPU utilization as part of its health monitoring metrics, even for burstable instance types like the t3a family, it's important to understand that the overall health assessment is more complex than just CPU utilization alone.

Elastic Beanstalk's enhanced health monitoring takes into account various factors to determine the health of your environment. CPU utilization is indeed one of these factors, but it's not the only one. The system also considers other metrics such as request throughput, latency, and HTTP status codes.

For burstable instance types, while they can handle short bursts of high CPU usage, sustained high CPU utilization over long periods is still considered a potential issue. This is because it could indicate that the instance is consistently exceeding its baseline performance and may not be adequately sized for the workload.

Regarding the specific thresholds for environment health component metrics, AWS does not publicly disclose the exact values. These thresholds are part of Elastic Beanstalk's internal algorithms and may be adjusted by AWS over time to improve the service.

However, you can view detailed health information and metrics in the Elastic Beanstalk console. The Health page in the environment management console displays health status, metrics, and causes for the environment and for each Amazon EC2 instance. This includes CPU utilization, along with other metrics like load average and request information.

If you find that the health status doesn't accurately reflect your application's performance, you might consider creating custom CloudWatch metrics or alarms that better suit your specific use case. This would allow you to define thresholds that are more appropriate for your burstable instances and workload patterns.

Remember, while CPU utilization is an important metric, Elastic Beanstalk's health monitoring is designed to provide a holistic view of your application's health, taking into account multiple factors beyond just CPU usage.
Sources
Monitoring environment health in the AWS management console - AWS Elastic Beanstalk
Enhanced health monitoring with the environment management console - AWS Elastic Beanstalk

profile picture
answered a month ago
profile picture
EXPERT
reviewed a month ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions