"A Parallel and Multithreaded ERLANG Implementation" - status?
Tue Jul 30 18:22:12 CEST 2002
Thank you Pekka for the update!
>Basically shortcut the distribution mechanism, detect Erlang nodes
>running on the same physical machine and then bypass the normal socket
>based distribution by establishing a shared memory communication area
>between the two Erlang nodes on the same machine. Far easier to
>implement and you avoid the overhead of socket, TCP and external format
>conversions. The full process contex switch overhead is still there and
>one also would need to handle "node" scheduling of Erlang processes for
>full benefit etc, etc. But thats the same problems as today with
How much that would enhance performance? The pipe-communication is supposed
to provide 10% improvement over TCP/IP (as
http://erlang.stacken.kth.se/user.html#unixdom-0.1 describes. Perhaps there
are newer and faster version somewhere?) so could one expect whole lot of
>A real MT Erlang is still desirable and it can be done very nicely and
>efficient in my oppinion. MP machines is, as mentioned, a comodity
>nowdays and now when hyperthreading is introduced in the Intel P4Xeon
>CPUs even more more interesting benefits for an MT Erlang is opened.
Letting the operating system do the balancing in one physical machine would
be very nice because that would leave out the own balancing code - which in
turn would give some performance advantage in larger SMP systems.
I sure hope, that MT Erlang would become under active development (or at
least some evaluation) in near future, to ensure that Erlang continue to be
interesting language for problems suitable for SMP-machines.
Student. . . . . . . . . .University of Oulu, Dept. of EE
Assistant Researcher . . .Apricot Project
"Computing is a field which has one of the shortest collective memories
of any engineering or scientific discipline." - Marty Fouts,
More information about the erlang-questions