14 articles
Breaking free from 'just add more resources.' From Amdahl's Law, USL, and queuing theory equations to practical capacity planning based on empirical analysis.
Stop tuning by guesswork. A cheat sheet on the three legendary Linux observability tools (perf, Ftrace, BPF) for safe and surgical performance analysis.
Moving away from guesswork tuning. From crisis tool prep to /proc secrets, kprobes/uprobes, and PMCs—here's how performance tools really work.
Don't be fooled by IOPS or average latency! A guide to virtual disk traps, bimodal distributions, and hunting mysterious I/O with biostacks.
Is the disk actually slow? A cheat sheet on logical vs physical I/O gap, cache traps, VFS latency, and benchmark pitfalls.
A review and takeaways from 'Team Geek'. From HRT principles to dealing with poisonous people, here is the mindset needed to build great engineering teams.
Graduating from 'just run tcpdump'. A cheat sheet covering TTFB traps, TCP backlog drops, TIME_WAIT port exhaustion, and tracking down true network latency.
Why allocated isn't consumed. A cheat sheet on demand paging, the horrors of swapping, allocators, and profiling real memory performance.
What happens behind 100% CPU usage? A skim-friendly cheat sheet covering run queues, context switching, IPC, and profiling tricks.
Why are system calls slow? What is %si? A skim-friendly cheat sheet for essential OS and kernel internals to survive performance analysis.
Diagnose elusive cloud performance traps, Noisy Neighbor effects, and container CPU throttling with this Chapter 11 skim-friendly cheat sheet.
Benchmarking results are famously misleading. Here are the critical traps you must avoid when evaluating system changes.
