You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@activemq.apache.org by Wolfgang Schmid <wo...@atos.net> on 2013/10/25 15:00:58 UTC

Deadlock in Broker (using SSL)

Hi,

we are running ActiveMQ 5.8 on a Solaris platform (Solaris 10) and we are
using SSL. That's the way how we defined the SSL stuff:

<broker ...>

        <transportConnectors>
          <transportConnector name="ssl"
uri="ssl://0.0.0.0:61617?needClientAuth=false&amp;maximumConnections=1000&amp;wireformat.maxFrameSize=104857600&amp;wireFormat.maxInactivityDuration=0&amp;transport.connectionTimeout=90000&amp;wireFormat.maxInactivityDurationInitalDelay=90000"
/>
        </transportConnectors>
       
        <sslContext>
         <sslContext keyStore="file:...../server-keystore"
                     keyStorePassword="some password"
                     trustStore="file:..../client-truststore"
                     trustStorePassword="some password"/>
        </sslContext>

</broker>


So currently we running tests without InactivityMonitor. (but we have the
same behavior if InactivityMonitor is active).

In our test scenario we have many clients that send messages (both to queues
and topics) and one server that processes these messages.
The broker is activated using the Tanuki service wrapper.

As you can see in the attached logfile of the service wrapper you can see
two situation, where the service warpper detects a deadlock in the broker
and therefore restarts the broker.

Any ideas why there is a dead lock ?
Why do I see a TCp connection even if we use SSL ?
INFO   | jvm 8    | 2013/10/25 06:54:13 | WrapperManager Error: "ActiveMQ
Connection Dispatcher:tcp://10.93.25.11:63414" tid=0

Any comments are appreciated.


wrapper.txt
<http://activemq.2283324.n4.nabble.com/file/n4673240/wrapper.txt>  








--
View this message in context: http://activemq.2283324.n4.nabble.com/Deadlock-in-Broker-using-SSL-tp4673240.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

Re: Deadlock in Broker (using SSL)

Posted by Christian Posta <ch...@gmail.com>.
Looks like you had some interesting SSL errors that caused the socket to
want to close.. and it must have raced against an existing write that was
in progress and also wanted to close...

for the logging of the tcp:// that's wrong.. when we retrieve the remote
address, we're just defaulting to using the text "tcp://" ... can you open
a jira for that and we can change that?

Are you seeing other exceptions in the log file (maybe on the client side)?


On Fri, Oct 25, 2013 at 6:00 AM, Wolfgang Schmid
<wo...@atos.net>wrote:

> Hi,
>
> we are running ActiveMQ 5.8 on a Solaris platform (Solaris 10) and we are
> using SSL. That's the way how we defined the SSL stuff:
>
> <broker ...>
>
>         <transportConnectors>
>           <transportConnector name="ssl"
> uri="ssl://
> 0.0.0.0:61617?needClientAuth=false&amp;maximumConnections=1000&amp;wireformat.maxFrameSize=104857600&amp;wireFormat.maxInactivityDuration=0&amp;transport.connectionTimeout=90000&amp;wireFormat.maxInactivityDurationInitalDelay=90000
> "
> />
>         </transportConnectors>
>
>         <sslContext>
>          <sslContext keyStore="file:...../server-keystore"
>                      keyStorePassword="some password"
>                      trustStore="file:..../client-truststore"
>                      trustStorePassword="some password"/>
>         </sslContext>
>
> </broker>
>
>
> So currently we running tests without InactivityMonitor. (but we have the
> same behavior if InactivityMonitor is active).
>
> In our test scenario we have many clients that send messages (both to
> queues
> and topics) and one server that processes these messages.
> The broker is activated using the Tanuki service wrapper.
>
> As you can see in the attached logfile of the service wrapper you can see
> two situation, where the service warpper detects a deadlock in the broker
> and therefore restarts the broker.
>
> Any ideas why there is a dead lock ?
> Why do I see a TCp connection even if we use SSL ?
> INFO   | jvm 8    | 2013/10/25 06:54:13 | WrapperManager Error: "ActiveMQ
> Connection Dispatcher:tcp://10.93.25.11:63414" tid=0
>
> Any comments are appreciated.
>
>
> wrapper.txt
> <http://activemq.2283324.n4.nabble.com/file/n4673240/wrapper.txt>
>
>
>
>
>
>
>
>
> --
> View this message in context:
> http://activemq.2283324.n4.nabble.com/Deadlock-in-Broker-using-SSL-tp4673240.html
> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
>



-- 
*Christian Posta*
http://www.christianposta.com/blog
twitter: @christianposta