Strace - Swiss Army Knife to Trace, Analyze, RT Debug
Talkat Open Source Summit Europe 2023 · on Thursday, 2023-09-21· at 11:55· Room 0E-2 (Floor 0), Abandoibarra Etorb., 4, 48011 Bilbao, Bizkaia, Spain·
summary
strace is used to record exact system calls, timestamps and children so you can find missing config files, slow syscalls and reproduce timing-dependent bugs. Typical workflow: run with -f to follow children, write output with -o or -ff per-pid files, filter syscalls with -e open,read,write and capture precise times with -ttt and -T (kernel time). Examples: 15 ms serial read latency traced to FTDI latency timer (disable it), replay timing with usleep to trigger a race, startup drops from ~4 minutes to ~11 seconds.
AI-written summary · about
chapters (19)
- 0:00strace introduction
- 2:12Where to find strace documentation and resources
- 4:42Basic strace usage
- 7:36Timing and forking
- 10:08Attach to running
- 13:00File output handling
- 15:27Filter system calls
- 17:38Verbose output
- 19:24Binary data representation and hexadecimal output
- 20:04Trace specific calls
- 23:37Timestamps analysis
- 26:49Kernel time measurement
- 28:14Debugging delays
- 31:043D graphics debugging
- 34:19Serial interface debug
- 43:30IoT timestamps extraction
- 46:20Network socket analysis
- 50:57Network timing issues
- 53:25Q&A
related talks
45mStrace: Monitoring The Kernel-User-Space Conversation
01h 31mUnderstanding Linux with Strace
23mSecuring Your Software Delivery Chain with Process Auditing
37mKernel and userspace tracing with LTTng and friends
01h 06mHow to debug Linux multi threaded code
26mLTTng for full stack tracing
48mEfficient and Large-scale Infrastructure Monitoring With Tracing
47mWhat is that Process Doing?