Prometheus Operators in the Wild

[angelicagardner]

Info

Speaker(s): Bartłomiej Płotka & Mahmoud Amin, Google
Video URL: https://youtu.be/pMlv626cyGU?si=XtI16ZGWWOKMKHoA

Official Abstract/Summary

Managing and scaling the Prometheus collection can be a challenge! For instance, you might want to have a static number of collecting Prometheus-es, scale collections using the hashmode feature, or dynamically scale it based per node. All with its own complexities and nuances. Luckily, for Kubernetes, there are convenient OSS Prometheus operators that automate those deployment models! This talk will introduce you to two OSS operators that can help you manage Prometheus-es on Kubernetes: Prometheus Operator and GMP Operator. The audience will learn about the benefits and drawbacks. On top of that, Max and Bartek will show how to set up and use those to forward your metrics efficiently to the remote metric backend of your choice e.g. Thanos, Cortex or a cloud provider. By the end of this talk, you will have a good understanding of the trade-offs, so you can make an educated choice on what to use. Speakers will also explain the path to upstreaming functionality to one operator one day.