eBPF (Extended Berkeley Packet Filter) is a revolutionary technology that allows executing sandboxed programs within the Linux kernel, providing unprecedented access to low-level system events with minimal performance impact.
This talk explores the inner workings of eBPF and demonstrates how it can significantly improve observability, performance tuning, and security in real-world scenarios.
Practical demonstrations using a Minikube environment compare eBPF-powered load balancers with traditional ones, highlighting the differences in throughput, latency, and packet visibility.
The session covers tools like bpftrace, bcc, and libbpf, which simplify the development and use of eBPF programs, making it accessible for DevOps engineers, platform SREs, and backend developers.
Aditya Krishnan, a Software Engineer at Baton Systems, shares his expertise in backend development, cloud-native infrastructure, and DevOps methodologies, emphasizing the potential of eBPF in revolutionizing the way developers understand and manage production systems.
Generated using GPT-4o-mini.
Share
More Videos of our talks
Privacy Preserving Machine Learning
Best Practices for Highly Available Applications on Kubernetes