Re: How to run ActiveMq server from Java code.

2010-07-27 Thread Johan Edstrom
BrokerService broker = new BrokerService(); // configure the broker broker.addConnector("tcp://localhost:61616"); broker.setBrokerName("Broker1"); broker.setUseJmx(false); broker.start(); On Jul 27, 2010, at 2:51 PM, szeldon wrote: > > Hi, > > I'm working

How to run ActiveMq server from Java code.

2010-07-27 Thread szeldon
Hi, I'm working on an application that uses JMS among other transport technologies. There are TestNG tests for much of the code. I want to create unit tests for JMS invocations. For this, I need to start a JMS server inside my unit tests. How to achieve this using ActiveMq? I was looking for an a

Problem when trying Pure Master/Slave without authentication

2010-07-27 Thread jzhang
I am trying Pure Master/Slave without authentication for learning purpose. It appears that the slave can not connect to master. I have master 'bkr1' and slave 'bkr2'. The activemq.xml of 'bkr2' (the slave) looks like: The port 61613 is the one for master. I started m

Re: Selector and payload

2010-07-27 Thread cobrien
I don't think most providers read the JMS as exclusionary but anyway if you want to use only API provided by the JMS specifications then you can. The examples included with ActiveMQ provide examples on how to send messages using only the JMS standard API. -clark www.ttmsolutions.com Act

Re: 5.3.2 : Small question on transport binding

2010-07-27 Thread cobrien
Eric, On startup, the Broker does a reverse DNS lookup which is unpredictable across os versions especially with multihomed machines. Keep this in mind especially when making the assumption: "if "name" = "localhost" binding on localhost only => not visible for the rest of the network" -clark

Re: master/slave status command

2010-07-27 Thread cobrien
if slave=false always it sounds like a bug to me. a similar bug was reported, AMQ-2669, that looks like this one and is reported fixed in 5.4.0. https://issues.apache.org/activemq/browse/AMQ-2669 -clark www.ttmsolutions.com ActiveMQ reference guide at http://bit.ly/AMQRefGuide Andrew Wasil

ActiveMQ Session Task thread name

2010-07-27 Thread nervousbadger
Hi there, We are using ActiveMQ v5.3.1. We have multiple sessions per connection, each processing incoming messages. The problem we have is that each session dispatch thread has the same name - "ActiveMQ Session Task". We would like each thread to have a unique name to aid debugging using our ap

Re: When ActiveMQ does flush non persistent messages to disk

2010-07-27 Thread Gary Tully
The global systemUsage memory limit is the trigger for messages to be spooled to disk. Per destination memory limits can be used for flow control and to limit the destinations use of a cache, but pending messages are only spooled to disk when the system usage limits are reached. On 27 July 2010 06

Re: Having Duplicate message when using failover and networkConnector .

2010-07-27 Thread Gary Tully
can you package up your test case or ideally mash it into a junit test case along the lines of some of the tests in: http://svn.apache.org/viewvc/activemq/trunk/activemq-core/src/test/java/org/apache/activemq/usecases/ BrokerQueueNetworkWithDisconnectTest and NoDuplicateOnTopicNetworkTest may prov

VMCursor Exhaustion

2010-07-27 Thread tmurray
ActiveMQ version 5.3.2 and 5.4 Snapshot ActiveMQ is setup as a standalone broker. NODE_REQUEST topic is non-persistent. I have been getting an “Invalid acknowledgment” exception, after a process which was subscribed to a non-durable topic was killed by a SIGKILL (kill -9). It appears as thoug

Re: master/slave status command

2010-07-27 Thread Andrew Wasilczuk
Hi Clark, > Could you satisfy your requirements using advisory messages to listen for > the ActiveMQ.Advisory.MasterBroker notification to keep track of who is the > master broker? > > http://activemq.apache.org/advisory-message.html Thanks for your answer. Yeah, I guess this could work for

Re: Correct way to build a mesh of brokers

2010-07-27 Thread Gary Tully
for future reference, the ordering of elements in the broker sequence is alphabetical - before and the difference with 5.4 is that the xml configuration file is validated against the schema. It was always alphabetical, just never enforced. On 26 July 2010 19:31, ripienaar wrote: > > > > ripiena

5.3.2 : Small question on transport binding

2010-07-27 Thread Eric-AWL
Hi Would someone mind completing this mapping or correcting my mistakes please ? tcp://0.0.0.0:port binding on all interfaces :port tcp://127.0.0.1:portbinding on localhost only => not visible for the rest of the network tcp://aa.bb.cc.dd:port

Re: JMSPriority and activeMQ

2010-07-27 Thread James Strachan
On 27 July 2010 07:01, wrote: > Yes, I reckon they should be int. Maybe setPriority should throw an > exception if the input priority is out of range. I'd have thought we'd have to add this to the JMS TCK as its a change in behaviour. Maybe for now just log a warning? That won't have any TCK iss

Re: Correct way to build a mesh of brokers

2010-07-27 Thread Rob Davies
That ordering issue is a pita - appologies - 5.4 should be out next week cheers, Rob On 26 Jul 2010, at 19:31, ripienaar wrote: > > > > ripienaar wrote: >> >> The configs for 5.3.2 is not compatible with 5.4, I'd say that renders 5.4 >> useless and I cannot test. Given the state of the docu