Operating systems notes

OS for Engineers: Performance & Debugging

A methodical way to find which of the four resources is choking your system, and the exact tool that proves it. Every section is a drawn page beside a clear written explanation, about 18 minutes in all.

  • Chapter 36 of 37
  • About 18 minutes
  • 8 sections

Sign in with Google or GitHub and you land straight on this page. DSA comes with the all labs plan and every ChaiCode batch.

The first drawn page of OS for Engineers: Performance & Debugging: Four resources, one method
The chapter's first page: Four resources, one method

In this chapter

  • Four resources, one method
  • CPU: load average and where the time goes
  • Reading top and htop like a pro
  • Memory: is it really full?
  • Disk I/O: throughput is not latency
  • Context switches, interrupts & too many threads
  • Tracing and profiling: strace, perf, flame graphs
  • The debugging playbook

Next chapters

Read the whole chapter

Sign in and the chapter opens with every drawn page, and all 38 operating systems chapters beside it.

DSA comes with the all labs plan and every ChaiCode batch.