strace Wow Much Syscall by Brendan Gregg.
From the post:
I wouldn’t dare run strace(1) in production without seriously considering the consequences, and first trying the alternates. While it’s widely known (and continually rediscovered) that strace is an amazing tool, it’s much less known that it currently is – and always has been – dangerous.
strace is the system call tracer for Linux. It currently uses the arcane ptrace() (process trace) debugging interface, which operates in a violent manner: pausing the target process for each syscall so that the debugger can read state. And doing this twice: when the syscall begins, and when it ends.
With strace, this means pausing the target application for every syscall, twice, and context-switching between the application and strace. It’s like putting traffic metering lights on your application.
…
A great guide to strace, including a handy set of strace one-liners, references, “How To Learn strace,” and other goodies.
If you are interested in *nix internals and the potential of topic maps for the same, this is a great post.
I first saw this in post by Julia Evans.