All of this seems very reasonable... Except for the fact that also nprocs() 
is returning 1 after calling julia -p 2 :D

I call julia from terminal using *export PATH* in bash_profile, but the 
behavior is the same if I go in the directory and call *./Julia -p 2*

Plus, I think I saw n workers after a call at julia -p n in Julia 0.4 on 
another machine, but not sure about it...

(btw the doc really is misleading, maybe should be restated in more 
clearly, but it's a minor issue)

Il giorno martedì 19 aprile 2016 15:55:13 UTC+2, Andre Bieler ha scritto:
>
> As Greg said, the total number of workers is one less than the number of 
> available processes.
> There is always one master process (with id = 1, and not considered a 
> worker) and the remaining workers.
>
> So for *julia -p 3 *you will get one master process and two workers.
>
> The documentation may be misleading in this context, as it says:
> Starting with julia -p n provides n worker processes on the local 
> machine. 
>
> But *nworkers()* is actually n-1.
> However, if you do a *@everywhere*, the expression will be executed on 
> all n
> processes.
>

Reply via email to