[erlang-questions] profiling network apps: fprof vs sequential tracing

[Fred Hebert]

2010/9/17 Björn-Egil Dahlberg <egil@REDACTED>

>
> You can use eprof to find processes + code that produces bottlenecks. It is
> somewhat limited compared to fprof (no callgraph for instance) but is is
> much much faster (especially after the rewrite to r14) and will likely
> identify problematic intersections of processes and code.
>
>
> Regards,
> Björn-Egil
> Erlang/OTP


I'm not sure when the rewrite happened, but I just tested this on R14A
(confirmed by joelr on R14B) and it seems like there is a bug with
eprof:analyse/2. Whenever there is no option passed, the sort defaults to
{sort, time}. However, whenever an option is passed (either {sort, calls} or
{sort, time}), the output is sorted by the number of calls.