You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@activemq.apache.org by Rukus520 <Cr...@yahoo.com> on 2008/04/17 01:51:26 UTC
ActiveMQ Usage
Hi,
I'd like to handle broker failure by queueing messages I'm sending (for
transmission after reconnecting) instead of depending on a backup broker. I
can either write code to do the queueing myself OR use an embedded,
in-process broker to do the queueing for me. The in-process broker (A) can
forward messages to the remote broker (B) when broker B is available, and
queue messages using the File Based Cursor when broker B is unavailable.
In essense, I'm using the remote broker as my JMS provider, and I'm
considering using the in-process broker just to take advantage of ActiveMQ's
built in queueing mechanism.
Is this usage of ActiveMQ appropriate? Why or why not?
If this is indeed a poor approach, is there any way to handle queueing on
the message producer side without having to manually detect connection
failure, and implementing the queueing yourself?
Thanks in advance.
--
View this message in context: http://www.nabble.com/ActiveMQ-Usage-tp16736018s2354p16736018.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.
Re: ActiveMQ Usage
Posted by Rob Davies <ra...@gmail.com>.
On 17 Apr 2008, at 00:51, Rukus520 wrote:
>
> Hi,
>
> I'd like to handle broker failure by queueing messages I'm sending
> (for
> transmission after reconnecting) instead of depending on a backup
> broker. I
> can either write code to do the queueing myself OR use an embedded,
> in-process broker to do the queueing for me. The in-process broker
> (A) can
> forward messages to the remote broker (B) when broker B is
> available, and
> queue messages using the File Based Cursor when broker B is
> unavailable.
>
> In essense, I'm using the remote broker as my JMS provider, and I'm
> considering using the in-process broker just to take advantage of
> ActiveMQ's
> built in queueing mechanism.
>
> Is this usage of ActiveMQ appropriate? Why or why not?
> If this is indeed a poor approach, is there any way to handle
> queueing on
> the message producer side without having to manually detect connection
> failure, and implementing the queueing yourself?
>
> Thanks in advance.
>
>
> --
> View this message in context: http://www.nabble.com/ActiveMQ-Usage-tp16736018s2354p16736018.html
> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
>
What you've outlined is perfectly valid usage scenario - and something
we usually recommend where you want local client persistence, and the
client not to block while the network/broker are not available
cheers,
Rob
http://open.iona.com/ -Enterprise Open Integration
http://rajdavies.blogspot.com/