[erlang-questions] Erlang on supercomputers

Ulf Wiger (TN/EAB) <>
Fri Apr 20 09:00:19 CEST 2007


... but in the ERTS User's Guide, there's a chapter
called:

"How to implement an alternative carrier for the erlang distribution"

http://www.erlang.org/doc/doc-5.5.2/erts-5.5.2/doc/html/alt_dist.html#3 

with a runnable example using Unix Domain Sockets instead of 
TCP/IP.


BR,
Ulf W

> -----Original Message-----
> From:  
> [mailto:] On Behalf Of 
> Bengt Kleberg
> Sent: den 20 april 2007 07:18
> To: 
> Subject: Re: [erlang-questions] Erlang on supercomputers
> 
> On 2007-04-19 22:52, Chad Brewbaker wrote:
> > The problems with Bluegene/L are:
> > 
> > 1. The compute node kernel only allows user process per 
> CPU. I haven't 
> > looked into the Erlang source much, but all threading must 
> be "green".
> 
> it is possible to disable smp, you choose at compile time. 
> (actually you could compile with smp but not use it, but that 
> would be slower).
> 
> 
> > 2. Communication is done over MPI in C/C++/Fortran. There 
> is no API for 
> > doing TCP/IP communication I know of, but if you are 
> feeling lucky you 
> > can use un-documented low level communication API's.
> 
> erlang uses tcp/ip.
> 
> 
> bengt
> -- 
> Those were the days...
>     EPO guidelines 1978: "If the contribution to the known art resides
>     solely in a computer program then the subject matter is not
>     patentable in whatever manner it may be presented in the claims."
> _______________________________________________
> erlang-questions mailing list
> 
> http://www.erlang.org/mailman/listinfo/erlang-questions
> 




More information about the erlang-questions mailing list