[openamq-dev] Multi-threaded connection requests fail?

Elli Barasch comptonsw at comcast.net
Thu Nov 8 21:10:49 CET 2007


Here are traces from my app and the server.  Note that I removed the 
main thread connection_new call.  Only the thread
functions themselves make the connection_new requests.

Elli

--------------        Client   -------------

14:51:16: Thread OS0 starting

14:51:16: TC Thread 1 starting

14:51:16:    1> amq_client_agent/Connection      Initialise-Connection:
14:51:16:    1> amq_client_agent/Connection         (--) 
Connect                         ->Negotiate-Connection
14:51:16:    1> amq_client_agent/Connection                + 
Set-No-Connection-Error
14:51:16:    1> amq_client_agent/Connection                + 
Connect-To-Server
14:51:16: E: Thread TC1 could not connect to localhost
14:51:16: Only a single thread may register signals.
14:51:16: Only a single thread may register signals.
14:51:16:    2> amq_client_agent/Connection      Initialise-Connection:
14:51:16:    2> amq_client_agent/Connection         (--) 
Connect                         ->Negotiate-Connection
14:51:16:    2> amq_client_agent/Connection                + 
Set-No-Connection-Error
14:51:16:    2> amq_client_agent/Connection                + 
Connect-To-Server

--------------------------   Server  -------------------

2007-11-08 14:50:47: W: security warning - logins use default passwords
2007-11-08 14:50:47:    6> amq_server_agent/Master          
Initialise-Master:
2007-11-08 14:50:47:    6> amq_server_agent/Master             (--) 
Ok                              ->New-Connection
2007-11-08 14:50:47:    6> amq_server_agent/Master                    + 
Initialise-Server
2007-11-08 14:50:47: I: server binding to 127.0.1.1:5672
2007-11-08 14:50:47:    6> amq_server_agent/Master                    + 
Allow-Thread-To-Settle
2007-11-08 14:50:47: I: server starting in non-clustered mode
2007-11-08 14:50:48:    6> amq_server_agent/Master                    + 
Report-Server-Ready
2007-11-08 14:50:48: I: server ready for incoming AMQ connections
2007-11-08 14:50:48:    6> amq_server_agent/Master                    + 
Accept-Connection
2007-11-08 14:51:48: I: cnn=0 msg=0 mem=0K/4387K exc=8 que=0 csm=0 bnd=0 
adx=0 idx=8 map=0
2007-11-08 14:51:48: I: qcn=0 cnq=0
2007-11-08 14:51:53:    6> amq_server_agent/Master          
Initialise-Master:
2007-11-08 14:51:53:    6> amq_server_agent/Master             (--) 
Interrupt                       ->Shutting-Down
2007-11-08 14:51:53:    6> amq_server_agent/Master                    + 
Report-Server-Interrupted
2007-11-08 14:51:53: I: caught signal, shutting down
2007-11-08 14:51:53:    6> amq_server_agent/Master                    + 
Shut-Down-Active-Connections
2007-11-08 14:51:53:    6> amq_server_agent/Master          Shutting-Down:
2007-11-08 14:51:53:    6> amq_server_agent/Master             (--) 
Finished                        ->
2007-11-08 14:51:53:    6> amq_server_agent/Master                    + 
Shut-Down-Server
2007-11-08 14:51:53:    6> amq_server_agent/Master          <Destroyed>
2007-11-08 14:51:53: I: shutdown is complete






Martin Lucina wrote:
> On 8/11/2007, "Martin Sustrik" <sustrik at imatix.com> wrote:
>
>   
>>> I have been stepping through the code.  Does GDB get in the way?
>>>       
>> Probably not, unless you've been pausing at specific point in
>> amq_connection_new function for longer than 30 secs.
>>     
>
>
> It'll get in the way sooner than that, if you're using heartbeats.
>
> The default with MT builds is to use a heartbeat of 2 seconds, and if you
> miss out 4 heartbeats (i.e. 8 seconds) the server will drop the
> connection.
>
> -mato
> _______________________________________________
> openamq-dev mailing list
> openamq-dev at lists.openamq.org
> http://lists.openamq.org/mailman/listinfo/openamq-dev
>
>   

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openamq.org/pipermail/openamq-dev/attachments/20071108/445ed684/attachment.htm 


More information about the openamq-dev mailing list