You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@activemq.apache.org by Oleg Dulin <ol...@gmail.com> on 2013/07/19 18:26:21 UTC

Odd network of broker behavior with 5.8

I have two brokers set up on two servers. One is 10.194.2.1, the other 
is 10.194.2.2 .

One of them has this in its config file:

          <networkConnectors>
                         <networkConnector name="2.1"
                                  uri="static:(tcp://10.194.2.1:3000)"  
duplex="true" />
                 </networkConnectors>

I have clients connecting to these two brokers using "failover" schema, 
i.e. failover:(tcp://10.194.2.1:3000,tcp://10.194.2.2:3000)

The clients start and end through out the day. They are desktop 
applications for the most part.

What we are observing is very odd behavior with 10.194.2.1 (we think 
that's the one). What seems to happen is that after awhile it either 
stops accepting client connections, stops handlings the queues and 
topics. No exceptions are thrown, nothing out of ordinary, it just 
simply hangs in there when someone tries to establish a connection, and 
when publishing messages it accepts them but doesn't forward them to 
consumers.

I switched the setup temporarily to a single message broker. That 
alleviated the symptoms.

However, I need to move this back to a 2-broker setup ASAP.

Am I doing something wrong ? Would be better off using two connections 
as opposed to duplex ?

Any thoughts are greatly appreciated.



-- 
Regards,
Oleg Dulin
http://www.olegdulin.com



Re: Odd network of broker behavior with 5.8

Posted by Oleg Dulin <ol...@gmail.com>.
Yes, I am reconfiguring my environment and going to set up a test 
situation again.

I'll keep you posted, thanks!

On 2013-07-22 23:08:40 +0000, Christian Posta said:

> --089e0160c406d911ae04e221c15a
> 
> Content-Type: text/plain; charset=ISO-8859-1
> 
> 
> 
> can you take a stack trace of the broker when it gets into that state?
> 
> 
> 
> 
> 
> On Fri, Jul 19, 2013 at 12:26 PM, Oleg Dulin 
> <ol...@gmail.com> wrote:
> 
> 
> 
> > I have two brokers set up on two servers. One is 10.194.2.1, the other is
> 
> > 10.194.2.2 .
> 
> >
> 
> > One of them has this in its config file:
> 
> >
> 
> >          <networkConnectors>
> 
> >                         <networkConnector name="2.1"
> 
> >                                  
> uri="static:(tcp://10.194.2.1:**3000<http://10.194.2.1:3000>)"
> 
> >  duplex="true" />
> 
> >                 </networkConnectors>
> 
> >
> 
> > I have clients connecting to these two brokers using "failover" schema,
> 
> > i.e. failover:(tcp://10.194.2.1:**3000 <http://10.194.2.1:3000>,tcp://
> 
> > 10.194.2.2:3000)
> 
> >
> 
> > The clients start and end through out the day. They are desktop
> 
> > applications for the most part.
> 
> >
> 
> > What we are observing is very odd behavior with 10.194.2.1 (we think
> 
> > that's the one). What seems to happen is that after awhile it either stops
> 
> > accepting client connections, stops handlings the queues and topics. No
> 
> > exceptions are thrown, nothing out of ordinary, it just simply hangs in
> 
> > there when someone tries to establish a connection, and when publishing
> 
> > messages it accepts them but doesn't forward them to consumers.
> 
> >
> 
> > I switched the setup temporarily to a single message broker. That
> 
> > alleviated the symptoms.
> 
> >
> 
> > However, I need to move this back to a 2-broker setup ASAP.
> 
> >
> 
> > Am I doing something wrong ? Would be better off using two connections as
> 
> > opposed to duplex ?
> 
> >
> 
> > Any thoughts are greatly appreciated.
> 
> >
> 
> >
> 
> >
> 
> > --
> 
> > Regards,
> 
> > Oleg Dulin
> 
> > http://www.olegdulin.com
> 
> >
> 
> >
> 
> >
> 
> 
> 
> 
> 
> -- 
> 
> *Christian Posta*
> 
> http://www.christianposta.com/blog
> 
> twitter: @christianposta
> 
> 
> 
> --089e0160c406d911ae04e221c15a--
> 
> 
> 


-- 
Regards,
Oleg Dulin
http://www.olegdulin.com



Re: Odd network of broker behavior with 5.8

Posted by Christian Posta <ch...@gmail.com>.
can you take a stack trace of the broker when it gets into that state?


On Fri, Jul 19, 2013 at 12:26 PM, Oleg Dulin <ol...@gmail.com> wrote:

> I have two brokers set up on two servers. One is 10.194.2.1, the other is
> 10.194.2.2 .
>
> One of them has this in its config file:
>
>          <networkConnectors>
>                         <networkConnector name="2.1"
>                                  uri="static:(tcp://10.194.2.1:**3000<http://10.194.2.1:3000>)"
>  duplex="true" />
>                 </networkConnectors>
>
> I have clients connecting to these two brokers using "failover" schema,
> i.e. failover:(tcp://10.194.2.1:**3000 <http://10.194.2.1:3000>,tcp://
> 10.194.2.2:3000)
>
> The clients start and end through out the day. They are desktop
> applications for the most part.
>
> What we are observing is very odd behavior with 10.194.2.1 (we think
> that's the one). What seems to happen is that after awhile it either stops
> accepting client connections, stops handlings the queues and topics. No
> exceptions are thrown, nothing out of ordinary, it just simply hangs in
> there when someone tries to establish a connection, and when publishing
> messages it accepts them but doesn't forward them to consumers.
>
> I switched the setup temporarily to a single message broker. That
> alleviated the symptoms.
>
> However, I need to move this back to a 2-broker setup ASAP.
>
> Am I doing something wrong ? Would be better off using two connections as
> opposed to duplex ?
>
> Any thoughts are greatly appreciated.
>
>
>
> --
> Regards,
> Oleg Dulin
> http://www.olegdulin.com
>
>
>


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