In this webinar, Grafana Labs Senior Engineering Manager Matvey Kukuy walks you through initial Grafana OnCall configuration, setting up an on-call management process, grouping, alert rendering, and organizing runbooks.

Welcome to the first episode of Site Reliability Chats with your hosts Jason Yee @gitbisect and Julie Gunderson @julie_gund.

As serverless architectures start to grow, finding the right troubleshooting approach becomes a business-critical aspect. This talk will dive into "the instrumentation approach" - keep track of internal events on the lambda and export processed telemetry data. We should handle legacy code, multiple code owners, and a massive stack of serverless technologies as with any real-life project. Our goal is to write as least code as possible, avoid any existing code changes, be cross-runtime, and leave no latency impact.

Serverless architecture poses a new challenge to developers - COST. We will discuss ways to find and optimize your application cost in this session.

Serverless has changed the way we build software for the better. But it’s also a paradigm shift that challenges many of our pre-existing practices and habits, like how we test our code and how we monitor their health. In this session, Yan Cui will share many of the lessons he has learned from running serverless in production over the last five years. Including tips on testing, observability, and how to keep your AWS cost in check.

***Please note that the audio turns on at the 45 second mark*** How Lumigo Monitors Its Production Monitoring your production is challenging. Lumigo processes more than 1.5B Lambda invocations per month and digests 25TB of data monthly. Many things can go wrong. In this session, I will explain how Lumigo operates its production from tools to processes.

When bugs inevitably make their way into production firmware, defensive programming practices give you peace of mind that your devices will keep running despite issues occurring in the background. However, defensive programming can lead to more issues than it solves if not used carefully. Often the better approach to take is to use the practice of offensive programming, which will help surface issues more quickly and easily and help you iterate towards bug-free firmware and a better overall product experience.

## Follow Cribl

Welcome to Shipa's Office Hours.

## Follow Cribl