[erlang-questions] Erl getting stuck in SMP with 8 cores

Sargun Dhillon xbmodder@REDACTED
Mon Sep 1 20:57:16 CEST 2008


This sounds almost like a hardware/kernel issue. I've run erlang on
8-CPU machines (32 & 64) without any "thread jamming." I was using
yaws & my own erlang code, both of which use recursive functions. The
difference was that I was running it under a fairly optimized Gentoo
system, and not CentOS. If you enable core dumps, do you get anything
from beam? When it locks up try SIGSEVC and get a core dump to
analyze. You can try (l/s)trace also in order to find system calls
which can be crashing on your box.


On Mon, Sep 1, 2008 at 10:09 AM, Kostis Sagonas <kostis@REDACTED> wrote:
> Eranga Udesh wrote:
>> Good to know that Erl works successfully wh 8 or more intel cores. Do
>> you use 8 schedulers with threads/hipe enabled?
>
> I've routinely used Erlang with SMP & HiPE (in fact running native code)
> with both 8 and 16 x86_64 cores for more than 6 months now without any
> problems.  No threads enabled though.
>
> The 'application' is running the native code compiler & dialyzer in
> parallel with as many schedulers as cores in the machine.
>
> Kostis
> _______________________________________________
> erlang-questions mailing list
> erlang-questions@REDACTED
> http://www.erlang.org/mailman/listinfo/erlang-questions
>



More information about the erlang-questions mailing list