[rabbitmq-discuss] Strange behavior found under Mac OS X

[Andres Rodriguez]

Additional info, Rabbitmq logs:

=CRASH REPORT==== 19-Nov-2008::17:31:58 ===
   crasher:
     pid: <0.217.0>
     registered_name: []
     exception error: bad argument in an arithmetic expression
       in function  rabbit_load:pick/0
       in call from rabbit_reader:compute_redirects/1
       in call from rabbit_reader:handle_method0/2
       in call from rabbit_reader:handle_method0/3
       in call from rabbit_reader:handle_input/3
       in call from rabbit_reader:mainloop/3
       in call from rabbit_reader:start_connection/3
     initial call: rabbit_reader:init(<0.185.0>)
     ancestors: [rabbit_tcp_client_sup,rabbit_sup,<0.81.0>]
     messages: []
     links: [#Port<0.277>,<0.185.0>]
     dictionary: []
     trap_exit: true
     status: running
     heap_size: 987
     stack_size: 23
     reductions: 706
   neighbours:

=SUPERVISOR REPORT==== 19-Nov-2008::17:31:58 ===
      Supervisor: {local,rabbit_tcp_client_sup}
      Context:    child_terminated
      Reason:     badarith
      Offender:   [{pid,<0.217.0>},
                   {name,tcp_client},
                   {mfa,{rabbit_reader,start_link,[]}},
                   {restart_type,temporary},
                   {shutdown,brutal_kill},
                   {child_type,worker}]


=INFO REPORT==== 19-Nov-2008::17:28:01 ===
accepted TCP connection on 0.0.0.0:5672 from 127.0.0.1:56603

=INFO REPORT==== 19-Nov-2008::17:28:01 ===
starting TCP connection <0.212.0> from 127.0.0.1:56603

=ERROR REPORT==== 19-Nov-2008::17:28:01 ===
Error in process <0.195.0> on node 'rabbit at andres-rodriguezs-macbook'  
with exit value: {{badmatch,{error,{fread,float}}}, 
[{cpu_sup,get_uint32_measurement,2},{cpu_sup,measurement_server_loop, 
1}]}


=INFO REPORT==== 19-Nov-2008::17:28:06 ===
closing TCP connection <0.212.0> from 127.0.0.1:56603



On Nov 23, 2008, at 10:54 PM, Andres Rodriguez wrote:

> Hi, I found something pretty interesting when trying to run any kind  
> of AMQP client under Max OS X. I tried with java, python and ruby  
> clients and none of them were working.
>
> Suspecting it could be my Rabbitmq or Erlang installs I tried  
> different setups:
>
> * Erlang from source and macports
> * Rabbitmq from source, tarball and macports
>
> Same result with any combination.
>
> With the help of tmm1 (AMQP client ruby gem creator) we discovered  
> that in line 67 of rabbitmq-server:
>
>    -os_mon start_cpu_sup true \
>
> Changing start_cpu_sup to false:
>
>    -os_mon start_cpu_sup false \
>
> made everything to work.
>
> Now my Rabbitmq broker is running pretty nice.
>
> Since I'm a newbie in these Erlang topics I wanted to share this  
> info and would like to know if someone expert at this subject could  
> identify an incompatibility in my machine's hardware since I already  
> found out that this cpu_sup thing is related to the processor (http://www.erlang.org/doc/man/cpu_sup.html 
> ).
>
> FWIW, I have a white macbook with a 2.4GHz Core 2 Duo (T8300).
>
> Thanks a lot!
>
> Andres.