Case study on CloudWatch custom metrics
The applications in your organization have something in common. They have all been built in the Java programming language running on the JVM runtime. The marketing team in your organization did a campaign that led to a huge spike of traffic in the application. This led to the application becoming very unstable and having frequent downtimes. You have gone through the Amazon CloudWatch automated dashboard for EC2 instances. The automated CloudWatch dashboard collects a lot of relevant information about the EC2 instances but still does not provide any insight into behavior due to the absence of a certain important EC2 metric. What is this metric? How do we go about collecting this metric to help us figure out what exactly is wrong with the application?
Solution
What is peculiar about Java applications and JVM is memory consumption. This can be figured out by installing the unified CloudWatch agent. This agent can help to collect memory-related...