[
https://issues.apache.org/jira/browse/ARTEMIS-2398?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16873303#comment-16873303
]
Justin Bertram commented on ARTEMIS-2398:
-----------------------------------------
bq. When connecting to a artemis cluster you need to supply a group address and
group port. I'm guessing that the machine that is running on that address and
the software running on that port acts as a sort of loadbalancer for all the
nodes?
That's sort of how it works. You *can* supply a group address/port to your
client in order to connect to a cluster, but you don't have to. You can supply
the client with the IP address/port for any node in the cluster, a full list of
IP addresses/ports for all the nodes in the cluster, or anything in between.
When using a group address/port you could end up connecting to any of the
cluster nodes because all of the cluster nodes should be broadcasting their
connector information on that same group address/port. Once you connect to a
particular cluster node the messages that the client sends will be
load-balanced according to the {{message-load-balancing}} configuration of the
{{cluster-connection}}.
In any event, it is possible to get a list of cluster nodes using the
{{listNetworkTopology}} operation on the {{ActiveMQServerControl}}. I recommend
the following:
* Read through the [management
documentation|http://activemq.apache.org/components/artemis/documentation/latest/management.html].
* Check out the relevant examples:
** {{management}} (in {{<ARTEMIS_HOME>/examples/features/standard/management}})
for a demonstration of using messages for management
** {{jmx}} (in {{<ARTEMIS_HOME>/examples/features/standard/jmx}}) for a
demonstration of using JMX for management
> In cluster mode, Artemis ignores createQueue command
> ----------------------------------------------------
>
> Key: ARTEMIS-2398
> URL: https://issues.apache.org/jira/browse/ARTEMIS-2398
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Reporter: Martijn
> Priority: Major
>
> We are using standalong artemis on our servers and this works great. We have
> created an autoconfigurer who automatically creates queues on startup. We can
> see that the queues are created.
> We are switching to a artemis cluster. We have updated our code and we can
> verify that our application is connecting to the new artemis cluster however
> the queues aren't created. It seams as if the createqueue command isn't
> passed on to the artemis nodes.
> To create the queues we use the following code:
> {code:java}
> // ManagementHelper.putOperationInvocation(message,
> "broker",
> "createQueue",
> queueName,
> queueName,
> null,
> true,
> "ANYCAST");
> {code}
> This works for standalone however not for a artemis cluster.
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
