Prometheus is the mainstream monitoring tool created for monitoring highly dynamic container environments. However it can still be used to monitor traditional container-less infrastructures. At its core Prometheus has the Prometheus Server that is composed of three components. A time series database that stores all the monitored metrics. A data retrieval worker responsible for pulling these metrics from target resources and saving them in the databse. And finally a HTTP web server that can receive PromQL queries to display metrics data using the Prometheus Web UI or other external applications (i.e., Grafana).
Categories: Big Data, Business Intelligence, Containers Orchestration | Tags: EC2, Druid, CNCF, Container Orchestration, Data Analytics, Helm, Metrics, OLAP, Operation, Cloud, Kubernetes, Prometheus, Python
Apache Druid is an open-source analytics data store which could leverage the auto-scaling abilities of Kubernetes due to its distributed nature and its reliance on memory. I was inspired by the talk…
Jul 16, 2019
Monitoring a production grade Hadoop cluster is a real challenge and needs to be constantly evolving. The software we use today is based on Nagios. Very efficient when it comes to the simplest…
Dec 21, 2018