Prometheus is a leading open source tool for monitoring and alerting. Originally, it was developed at SoundCloud and was later open sourced. It is also part of Cloud Native Computing Foundation (CNCF), like KubPrometheus server: This is the heart of Prometheus and does the main job of scraping and storing time series data. pull-metrics pulls the time series data from jobs or exporters using HTTP endpoints. For example, Prometheus would pull the different metrics and other data from api-service based on the configuration done on Prometheus (.yml file). Once metrics are scraped, they are stored locally. Then, Prometheus server runs rules on the stored data to generate time series data or raise alerts.

Pushgateway: This is used for fetching data from short-lived jobs. Short-lived jobs push metrics before exiting to Pushgateway. Prometheus server pulls metrics data from Pushgateway for scraping.

Alertmanager: This component is responsible for alerting. Prometheus server pushes raised alerts to Alertmanager, which sends them to different clients, such as email and so on.

PromQL: PromQL is a functional expression language or a query language. It allows you to select and aggregate data in real time. You can then consume this data to create meaningful information such as charts, tables, and so on, for example, on Prometheus Web UI. We'll consume PromQL in another UI tool, Grafana.ernetes.

https://prometheus.io/