We are always looking for speakers for our next events. If you want to talk please contact us or simply submit a talk here: https://forms.gle/e9rcrJEF7FK7J8o8A
Scaling Prometheus at Trade Republic: The Hard Way
At Trade Republic, rapid company growth is pushing our current Prometheus deployment – a single instance per environment deployed in a high-availability setup – to its limits. In this talk, I'll share the practical challenges, common pitfalls, and key lessons learned as we prepare to scale Prometheus at the collection layer. We'll explore Prometheus's internal constraints, such as resource consumption driven by cardinality and active time series, and discuss our current sharding strategies. Our approaches have their pros and cons but fit well into our use case – though they're not silver bullets for every scenario.
Break
kubezonnet: Monitor Cross-Zone Network Traffic in Kubernetes
Kubezonnet, an eBPF and Prometheus tool that reveals cross-zone (Kubernetes) network traffic. This talk covers its architecture and practical use cases for optimizing cost. We will briefly look at how we improved our Thanos setup by cutting costs across zones.
