Performance Wins with BPF: Getting Started
summary
Installs the BCC and bpftrace toolsets, then runs lightweight probes like execsnoop, opensnoop, tcplife and biosnoop for fast wins. Example: iostat shows 10–20% disk use but BPF attributes heavy reads to a runaway Perl log-rotator launching every second, fixed by killing the process. Recommends using libbpf/BTF-compiled agents (single ~150KB relocatable BPF binaries) for cross-kernel deployment and building GUIs that invoke these tools rather than rewriting core BPF utilities.
AI-written summary · about
chapters (14)
- 0:00Performance with BPF intro
- 1:05What is BPF
- 2:27BPF docs challenges
- 4:13Think like sysadmin
- 5:57Top 5 BPF tools
- 7:43Cassandra perf case study
- 9:10BPF observability wins
- 10:56BPF future dashboards
- 13:24Building BPF tools
- 15:09When to code BPF
- 16:56BPF Trace example
- 18:21BPF libbpf and BTF
- 19:23BPF kernel vs user mode
- 20:26BPF key takeaways
related talks
35mPerformance Insights into eBPF Step by Step
36mObservability and performance analysis with BPF
47mDive into BPF performance tools using python
32mBPF: Introduction, Programming Tips and Tricks
46mbpftrace internals
30mPerformance insights into eBPF step by step
25mKernel tracing with BPF
28mbpftrace: what is it, what's new, and where is it going