Grafana & Friends Munich

Welcome to our summer Meetup of 2025, hosted by the lovely folks of QAWare.

Talk 1)
Rethinking Cloud Observability – An Approach with Grafana Beyla and Pyroscope
Franz Wimmer, QAWare

Traditional monitoring approaches reach their limits in modern cloud environments. Distributed systems, dynamic scaling, and complex dependencies require new methods for monitoring and optimizing applications. Cloud observability is based on logs, metrics, and traces, which together provide a comprehensive picture of system health.Grafana Beyla and Grafana Pyroscope are powerful open-source tools available for in-depth analysis. Beyla uses eBPF for agentless tracing and metrics collection with minimal overhead, while Pyroscope enables continuous profiling to uncover bottlenecks at the code level. This combination allows developers to specifically identify and resolve latency, resource usage, and performance issues.The targeted use of these technologies helps optimize cloud architectures, reduce downtime, and sustainably improve system performance.Both solutions can be seamlessly integrated into Kubernetes, the industry standard for cluster orchestration. We’ll see how this works in practice through a live demo.

Talk 2)
Putting the Pro in Profiles
Ismail Simsek, Grafana Labs
Continuous Profiling is the next frontier in observability, promising deep insights into system performance with minimal overhead. Yet for many developers, the thought of flamegraphs, stacks, and CPU cycles can feel like black magic. In this talk, we'll break down profiling into something not just accessible, but empowering.
You'll discover why profiles are a missing pillar of observability, how continuous profiling changes the game compared to traditional tools, and why flamegraphs — despite their intimidating looks — are actually intuitive and powerful once demystified. Along the way, we'll explore practical examples where profiling turned sluggish systems into high-performers and revealed optimization opportunities worth millions.
Key takeaways:

• Understand how to use flamegraphs to spot and fix performance bottlenecks fast
• Learn practical techniques to optimize CPU, memory, and network resource usage
• See how continuous profiling transforms system observability and cost efficiency
• Gain confidence to apply profiling insights in production environments at scale

By the end of this talk,
you’ll be able to decode flamegraphs with confidence, uncover hidden performance issues across your stack, and apply continuous profiling to drive real improvements

Talk 3)
AI at Grafana
Maurice Rochau, Grafana Labs
This talk explores our AI offerings in Grafana Cloud and how we bring AI into observability. We'll introduce you to forecasting and outlier detection, must have capabilities in every observability stack, and Sift, our diagnostic assistant. We'll end by giving you an exclusive sneak peek into our newest addition, the Grafana Assistant