Prometheus: The Documentary (Category Management )
Interesting half-hour documentary The appearance of the Prometheus project in 2012 SoundCloud has become the de facto standard for cloud-native applications.
The story went something like this.
- Two ex-googlers. (Julius Volz, Björn Rabenstein) Apply to the startup SoundCloud and are responsible for the reliability of this solution. SoundCloud engineers had already made their own workload orchestrator. (That was before the K8s came along.) It is very difficult to monitor what is happening inside this cluster using statsd and graphite, and two ex-googlers first try to find an acceptable open-source solution, and then in their spare time they start doing a monitoring system like Google used to do for the Borg orchestrator. This leads to a plan to create a new monitoring system with a pull approach, a time-series base, and a user-friendly language for querying over. (PromQL)
- They're good at it, and they're going to promote it inside, and it's getting a good reception. From the very beginning, the code is published openly on GitHubBut then guys. announce This system is out on behalf of SoundCloud and another company that was an early adopter. Around this time, Kubernetes becomes opensource, and the Cloud Native Computing Foundation appears around it. (CNCF) The guys at SoundCloud communicate with CNCF and then Prometheus is transferred to this foundation and becomes the de facto standard in the K8s world. This close knot of K8s and Prometheus works well together and drives both projects forward. As a result, everything goes according to plan and now we have a de facto standard for monitoring:)
I’ve talked about other documentaries in the same series:
#Kubernetes #Film #Documentary #Software #Architecture #DistributedSystems #SRE