Pre-requisites of Monitoring System

Monitoring: metrics and alerting

A good monitoring system needs to clearly define what to measure and in what units (metrics). The monitoring system also needs to define threshold values of all metrics and the ability to inform appropriate stakeholders (alerts) when values are out of acceptable ranges. Knowing the state of our infrastructure and systems ensures service stability. The support team can respond to issues more quickly and confidently if they have access to information on the health and performance of the deployments. Monitoring systems that collect measurements, show data, and send warnings when something appears wrong are helpful for the support team.

To further understand metrics, alerts, and their connection with monitoring, we’ll go over their significance, potential benefits, and data we might want to keep track of.

Metrics

Metrics objectively define what should we measure and what units will be appropriate. Values of metrics provide an insight into the system at any point in time. For example, a web server’s ability to handle a certain amount of traffic per second or its ability to join a pool of web servers are examples of high-level data correlated with a component’s specific purpose or activity. Another example can be measuring network performance in terms of throughput (Mbps) and latency (round-trip time). We need to collect values of ...