In today's DevOps world, application deployments are proactively monitored from day one. Not only can you monitor the workloads but also the underlying infrastructure and how they impact your business. For example, if you are an e-commerce store, you deploy a web application and few microservices to a Kubernetes cluster. As part of the technology and processes you expose business metrics in your applications. A business metric could be something like dollar figure of the revenue recorded at each minute. Monitoring systems fetch this business metric often and can alert you if it breaches a threshold. Similarly, you can also monitor whether a microservice is up and running and how well it is performing. You can also monitor the underlying infrastructure such as a availability of a database system.
Monitoring solves several problems:A unified monitoring stack centralizes alerting and dashboarding capabilities. Fetch metric data from your application and infrastructure into the unified stack. Declarative define alerting rules. Build auto-correct mechanisms. Our stack is based on open source software: Prometheus, Alert Manager, Grafana, StatsD. The stack integrates well with cloud provider services such as AWS CloudWatch.
You can also mix and match your monitoring with third-party platforms such as DataDog, NewRelic and Sentry.