[erlang-questions] poll: monitoring performance of production erlang nodes

[Joel Reymont]

Paul,

On Nov 2, 2008, at 9:18 PM, Paul Mineiro wrote:

> (We could use traces now to collect
> timing statistics if we were willing to make intermediate functions  
> for
> code regions of interest).

Why not trace and time message paths? One time stamps are enabled,  
you'll have "entry" and "exit" points for messages, function calls,  
etc. You could then collect data and use and external tool to analyze  
the paths for you.

--
wagerlabs.com