You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@activemq.apache.org by archa <ar...@persistent.co.in> on 2011/05/12 06:43:34 UTC
Native Clustering in ActiveMQ
We are using Activemq 5.2 with journaledJDBC using DB2 database. Is Native
Clustering possible with ActiveMQ ?
--
View this message in context: http://activemq.2283324.n4.nabble.com/Native-Clustering-in-ActiveMQ-tp3516488p3516488.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.
Re: Native Clustering in ActiveMQ
Posted by Johan Edstrom <se...@gmail.com>.
You want your master slave to use "master/slave"
then you tell your clients about the failover, also look at these pages:
http://activemq.apache.org/masterslave.html
http://activemq.apache.org/jdbc-master-slave.html (Since you mentioned DB2)
http://activemq.apache.org/shared-file-system-master-slave.html
http://activemq.apache.org/failover-transport-reference.html
Note the
This is new in version 5.4:
Down on the page as well.
Cheers!
On May 12, 2011, at 12:28 AM, archa wrote:
> Can you tell me where exactly to mention the failover syntax.
> Eg:
> failover:(tcp://primary:61616,tcp://secondary:61616)?randomize=false
>
> Should this go in the transportconnector tag ?
>
> --
> View this message in context: http://activemq.2283324.n4.nabble.com/Native-Clustering-in-ActiveMQ-tp3516488p3516585.html
> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
Re: Native Clustering in ActiveMQ
Posted by Johan Edstrom <se...@gmail.com>.
That goes in the clients connector.
On May 12, 2011, at 12:28 AM, archa wrote:
> Can you tell me where exactly to mention the failover syntax.
> Eg:
> failover:(tcp://primary:61616,tcp://secondary:61616)?randomize=false
>
> Should this go in the transportconnector tag ?
>
> --
> View this message in context: http://activemq.2283324.n4.nabble.com/Native-Clustering-in-ActiveMQ-tp3516488p3516585.html
> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
Re: Native Clustering in ActiveMQ
Posted by archa <ar...@persistent.co.in>.
Can you tell me where exactly to mention the failover syntax.
Eg:
failover:(tcp://primary:61616,tcp://secondary:61616)?randomize=false
Should this go in the transportconnector tag ?
--
View this message in context: http://activemq.2283324.n4.nabble.com/Native-Clustering-in-ActiveMQ-tp3516488p3516585.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.
Re: Native Clustering in ActiveMQ
Posted by Johan Edstrom <se...@gmail.com>.
Oh,
Yes, then you are on the right track.
ActiveMQ supports 3 models.
Memory (Don't use this)
Disk (Use this for speed, niceness and whatnot else)
JDBC (It'll be slow)
/je
On May 11, 2011, at 11:06 PM, archa wrote:
> My requirement is as follows:
>
> I have my application which currently write to a single broker queue.
> Incase the broker goes down, my messages are lost, till I restart the broker
> and the consumers again (as consumers are formed on application startup).
>
> To avoid the loss of messages, I want sort of a master-slave configuration,
> so that when the master goes down, the slave broker continues consuming the
> messages, without requiring the message producer to change the broker ip in
> its configuration.
>
> I read about master-slave configuration in
> http://activemq.apache.org/jdbc-master-slave.html . Here it mentions about
> failover, but how is the failover mentioned in the activemq.xml (cant see it
> in the example given in that url). Also, will the failover be same for the
> master and slave, as it is a cluster ?
>
> --
> View this message in context: http://activemq.2283324.n4.nabble.com/Native-Clustering-in-ActiveMQ-tp3516488p3516511.html
> Sent from the ActiveMQ - User mailing list archive at Nabble.com.
Re: Native Clustering in ActiveMQ
Posted by archa <ar...@persistent.co.in>.
My requirement is as follows:
I have my application which currently write to a single broker queue.
Incase the broker goes down, my messages are lost, till I restart the broker
and the consumers again (as consumers are formed on application startup).
To avoid the loss of messages, I want sort of a master-slave configuration,
so that when the master goes down, the slave broker continues consuming the
messages, without requiring the message producer to change the broker ip in
its configuration.
I read about master-slave configuration in
http://activemq.apache.org/jdbc-master-slave.html . Here it mentions about
failover, but how is the failover mentioned in the activemq.xml (cant see it
in the example given in that url). Also, will the failover be same for the
master and slave, as it is a cluster ?
--
View this message in context: http://activemq.2283324.n4.nabble.com/Native-Clustering-in-ActiveMQ-tp3516488p3516511.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.
Re: Native Clustering in ActiveMQ
Posted by Johan Edstrom <se...@gmail.com>.
Clustering so to say kinda sorta goes against the JMS
spec, the idea is to get messages from A->B as quickly as possible.
There are HA solutions for stability and you can use networks
of brokers to provide scalability but the idea of "clustering" messages,
i.e replicating messages stores is kinda sorta counter to the target use-case
of messaging?
Or did I completely miss what you are asking?
On May 11, 2011, at 10:43 PM, archa wrote:
> We are using Activemq 5.2 with journaledJDBC using DB2 database. Is Native
> Clustering possible with ActiveMQ ?
>
>
> --
> View this message in context: http://activemq.2283324.n4.nabble.com/Native-Clustering-in-ActiveMQ-tp3516488p3516488.html
> Sent from the ActiveMQ - User mailing list archive at Nabble.com.