Every production-grade system requires a monitoring service that gives administrators a bird's-eye view of the health of the system. The health of the system can be gauged by operational parameters, like memory utilization, CPU utilization, network throughput, and more. Any deviation from the normal values of these parameters needs to be caught and reported, as it has the potential to cause a massive disruption to the system.
For example, a spike in the memory utilization of a computer resource needs to be reported promptly, as its prolonged persistence can severely impair the system.
Monitoring a production system is a critical non-functional requirement that cannot be ignored.
In a non-serverless, partially managed infrastructure ...
