[rabbitmq-discuss] persistence

Wed May 21 15:44:46 BST 2008

Zitat von Matthias Radestock <matthias at lshift.net>:

> Lars Bachmann wrote:
>> the log messages i sent you, were created by the communication with  
>>  one client where the error occurs. However i made some more  
>> actions  and Im sending u the files now. The "sasl" files are  
>> always empty  but in the log files is one error message.
>
> I do not see any error messages in the log files. Also, there is still
> no trace of the TCP listener starting, or the client
> connecting/disconnecting, which is extremely weird.
>
> How exactly are you starting the broker?
>
> Are you running a cluster?
>
> One thing to try is to get an erlang shell in your broker and evaluate
>   rabbit_log:info("show me!").
> and then check whether the message appears in the rabbit log.
>
> It would also be useful if you ran a clean test. Stop the broker remove
> the mnesia dir and log files. Start the broker. Then run your tests.
> And then send me the log files.
>
>
> Matthias.

the error message is in the broker4.log.bak file and it says:

=ERROR REPORT==== 21-May-2008::13:22:48 ===
Error in process <0.166.0> on node 'broker4 at vmpontos' with exit value:  
{badarg,[{erlang,port_command,[#Port<0.258>,[<<7 bytes>>,<<4  
bytes>>,<<1  
byte>>]]},{rabbit_writer,internal_send_command_async,3},{rabbit_writer,handle_message,2},{rabbit_writer,mainloop,1}]}

and there are also messages about tcp connections (for example):

=INFO REPORT==== 21-May-2008::13:15:39 ===
started TCP listener on 10.1.2.130:5672

I run your test with log info and the text showed up in the logs and  
no, i dont run a cluster. I ll ran a clean test now and send u the  
result.

Lars