Hi
Failed to build body from bytes. Reason: java.io.IOException: No
ClassLoaders found for: com.corpus.biemedia.servlet.MailEx
I'd say you don't have the MailEx class on the same path as ActiveMQ.
I suspect it's not good enough to have it inside the webapp because
the activemq listener
We successfully use ActiveMQ on the Unix partition of the z/9. We only
run the client part there (the broker runs on a x86-linux box), but
I'd think that the server has any issues running.
You cannot use the MQ queues directly in ActiveMQ, but you can setup a
bridge to transfer the messages. Of
I consider the 100gb memory limits a bit excessive.. I don't know your
production system, but if you get 100gb memory per queue I absolutely
want to own it :) This values refer to available RAM, I think
something like 20mb should be enough for most purposes.
Depending on the environment you're
Hi Vadim
The sender will block until resumed means that if your broker dies
(or when your broker is down) while your application calls send (or
commit) then that call will just wait until your broker is back up or
the slave broker has taken over. This is normally desired behavior.
However it can
I'd say you run out of memory on the broker due to a memory leak in
the connection handling, every closed connection leaves a thread open
and still hogs some memory. You can check that with jConsole.
Can you try reusing or pooling the connections? There's a
PooledConnectionFactory that does this.
I've seen the same behaviour today with an ActiveMQ 5.1.0 (JDBC-only
storage) and Java consumers (transactional).
The cause in my case was the AMQ-1838 bug which resulted in a huge
number of messages being paged in and the fetcher from the queue had
it's AtomicLong-lastMessageId set to a number
The correct MBean name is:
org.apache.activemq:BrokerName=name,Type=Broker
The jconsole (at least with Java 6) displays the object name of any MBean.
Also it might be worth to look at the activemq_web project which
provides a layer to access the broker over jmx (query queues etc).
This is then
If you really want to create and list queues you could use the
JMX-interface to the broker and call the addQueue method on the
Broker-MBean. That does create an empty queue.
Mario
On Wed, Jun 25, 2008 at 3:20 PM, James Strachan
[EMAIL PROTECTED] wrote:
2008/5/8 shaf [EMAIL PROTECTED]:
Thanks
I think the problem might be the default Subscription Recovery Policy
in ActiveMQ 4.1.1 (support for retroactive consumers).
See https://issues.apache.org/activemq/browse/AMQ-1321 for the problem
and http://activemq.apache.org/subscription-recovery-policy.html for
more information. I'd probably
I'd guess that the topics with 0 received don't have any consumer, the
ones with x have a single consumer and the ones with 2x have two
consumers registered. The two consumers both receive the message, so
two messages are received.
Mario
On Tue, Jun 24, 2008 at 11:28 PM, Badri
[EMAIL PROTECTED]
Hi
Depending on the JDBC driver you can have the driver to auto-reconnect
(I think jTDS supports that). However there's a catch in that: The
broker looses the lock on the master/slave table, so you'll end up
with both being the master (if you use jdbc-master/slave). I didn't
check that with the
Your problem might be that the two brokers are in a network of brokers
due to this line
networkConnector uri=multicast://default/
in the configuration (discovery via multicast).
However your config for broker B shouldn't do multicast discovery, but still..
Else you'll need to use
Did you close the session associated with the consumer?
Mario
On 3/6/08, sparky2708 [EMAIL PROTECTED] wrote:
Periodically I have the following error. (I don't close the producer anywhere
in my code) Any ideas why?
javax.jms.IllegalStateException: The producer is closed
I don't know if it applies to your problem but you could setup a Camel
routing on the broker so that the 10'000 queues are forwarded into a
few, so that each client would only have to listen to 10 or so queues
(like one per broker).
Mario
On 2/19/08, Ben Chobot [EMAIL PROTECTED] wrote:
Hey
I'd say your problem is due to the session running in auto-ack mode
(so after the onMessage method completes) and your manual ack. I think
ActiveMQ might just ignore your call to acknowledge() and wait for the
method to complete. Since the 5 second sleep is inside the
onMessage-method it'll assume
Hi
We once had the same issue (also with 4.1.1) and then discovered that
the messages were assigned to some consumer that basically crashed but
was still connected to the broker. Since 1000 messages is the default
prefetch size for a queue this might also be your problem. In JMX you
can see the
I'd say it's a problem related to connection pooling (or the lack
thereof). To do a simple test try replacing your
ActiveMQConnectionFactory with the PooledActiveMQConnectionFactory
(not sure about the name, but something along this line)
Mario
On 12/10/07, andriy_heikal [EMAIL PROTECTED] wrote:
Just use the userName (uppercase N) and password property. This
applies to JNDI as well as spring.
Example tomcat config:
Resource
name=jms/ConnectionFactory
auth=Container
type=org.apache.activemq.ActiveMQConnectionFactory
description=JMS Connection Factory
I think that's a reasonable scenario, Oracle will actually have
trouble detecting them fast enough. I don't know if there's a setting
in Oracle that would speed up the checks. MySQL and MS SQL Server
detect the disconnect almost instantaneous.
However this should not happen on a clean shutdown.
Since Version 5 isn't released yet the downloads are found under
Latest SNAPSHOTS.
http://people.apache.org/repo/m2-snapshot-repository/org/apache/activemq/apache-activemq/5.0-SNAPSHOT/
and then the newest tar.gz or zip.
Mario
On 10/9/07, Epiphany [EMAIL PROTECTED] wrote:
However, when I click
I think you're issue is the prefetching of messages that is done by
ActiveMQ. Set the prefetchSize to 1 and it should work as you expect
it. See http://activemq.apache.org/what-is-the-prefetch-limit-for.html
Mario
On 10/2/07, Ned Wolpert [EMAIL PROTECTED] wrote:
Folks-
I'm trying to see how
Hi
If you're using ActiveMQs master-slave then a slave will never have
it's transport connector up. This is archived by a database lock or
file lock or by having the slave checking the master (pure master
slave). See http://activemq.apache.org/masterslave.html
So you can just use the failover
This works for me. As soon as the database detects the loss of the
connection it frees the lock. Tested with MySQL and SQL-Server. I
didn't plug the power but instead did a kill -9 or unplugged the
network cable on the master.
I think this is supposed to be a feature of any database or else you'd
You can specify your own jdbc adapter class (see
http://activemq.apache.org/jdbc-support.html).
What I'd do is write a mysql adapter (extend the mysql one) and
override the locking statements. You can also rename the mysql adapter
for AMQ 5 (or svn revision 518161/518164) into
Hi
We had the same problem with NFS, I think it's an NFS/Java issue since
there also seem to be problems with other applications.
But someone told me they've it running successful with NFS, so I
suppose it's NFS-implementation/version specific whether locking works
or not (he ran a JVM 1.5.0_08
(f.e. webconsole-properties.xml) and the system
property tells it which one to use.
Else your settings are looking good to me.
Mario
On 8/28/07, den!s [EMAIL PROTECTED] wrote:
Mario Siegenthaler-2 wrote:
Only if the two are a master/slave pair (only one running at a time).
This is what
On 8/24/07, JohnFish [EMAIL PROTECTED] wrote:
The question now is can I have a message producer in openjms and use the
ajax functionality of ActiveMQ to consume these messages? Thanks in advance.
I think you'll have to setup a JMS bridge with active-mq to get the
messages from openjms into
hi guys
What's the expected behavior with messages that are not ack'ed (for
example because the MessageListener throws an exception or the
transaction is rolled back) and have a JMSXGroupID set?
Say we've two messages A and B that belong to the same message group.
1) Processing of A fails. The
will try to embedded the activeMQ to Tomcat.
But for the moment, the session never be created but i have a connection.
I can't understand what happend because there is no exception throw !!
Could you help me again ?
Thanks in advance,
Regards,
denez
Mario Siegenthaler-2 wrote:
I think
On 8/9/07, keneida [EMAIL PROTECTED] wrote:
is it possible to set persistance to oracle in web console's emedded broker?
when i try to do it i get:
Yes it's possible, you can change the activemq.xml in the WEB-INF of
the war to whatever configuration you like.
Exception processing TLD
The default JDBCPersistenceAdapter has access to the BrokerService.
The broker service exposes the ManagementContext that contains a
getMBeanServer-method. So you get easy access to the MBean-Server.
Just register your JMX-Bean there. Example:
// just to illustrate..
Yes, this is very annoying in some situations where you depend onto
this numbers. How hard / expensive (performance wise) would it be to
add an configuration flag so the numbers are actually fetched from the
message store instead of 'just' being the difference of (messages
sent)-(messages
Hm, that's supposed to work, although I would create a
webconsole-custom.xml instead of changing the
webconsole-properties.xml, but that's minor.
We're running a configuration that's almost the same as yours and its
running just fine. Can you provide some more details about how you
start it? Do
I think you're supposed to call connection.start() as soon as you're
application thinks it fitting, connections don't come started (and I'd
hate it if they did). By documentation you can create session on
stopped connections you just won't receive any messages (not even sure
if you could send some
Just a thought, but why don't you use a topic for this? I think a
queue will cause problems because the message will be stored. So the
I'm here thing would possibly refer to an older question and would
no longer be relevant.
Hm, just reread your question, and realized that you directly want to
Yes it's possible, we use it in a production environment with ActiveMQ 4.1.1
Just follow the description on
http://activemq.apache.org/web-console.html under Starting the Web
Console in a seperate VM/in a Web-Container
Mario
On 7/25/07, kpetrov [EMAIL PROTECTED] wrote:
Since 5.0 is not
is present in
Tomcat/conf directory. Iam using active mq 4.0 is embedded broker
supported in this version ?
Any idea as to why the factory is becoming null ?
Thank you,
Suchitha.
-Original Message-
From: Mario Siegenthaler [mailto:[EMAIL PROTECTED]
Sent: Thursday, June 28, 2007 2:57 PM
Did you check the Sharing embedded broker across webapp
contexts-thread (last week)? This was the exact same problem and
serveral possible solutions (at least one of them I tested
personally).
Easy way: put the activemq in the common/lib and just let the to
webapps connect to vm://localhost. Ways
Somebody proposed in a different thread to use a MessageSelector to
only select the messages that are at least an hour old. You could set
a message-property to System.currentTimeMillis() and then create a
selector with SendTime +(System.currentTimeMillis()-360).
You'd have to recreate your
I've added some documentation about the webconsole and I included a
section about how to password protect it (although only on jetty, but
I hope the folks will figure out how to 'translate'' that to tomcat by
themself :).
And for the JMX-password issue there is a patch by Andrew Deason
We've tried it on NFS on linux and it didn't work (some locking issue
with java and nfs). However it worked with windows shares, but we're
not using it because we run linux boxes.
Mario
On 5/31/07, Michael Slattery [EMAIL PROTECTED] wrote:
Hello List!
I'm interested in putting together the
Yes that might work. If I understand it correctly you could specify on
the maven command line which war to build resp. which 'version' of it
to run?
As you may have realized I had quite some trouble in finding a way to
let the user/deployer specify which 'version' to run. The best thing I
could
Many thanks James, I realize this patch is quite bitchy to apply :)
I'll try it out as soon as I'm back home.
Mario
On 5/24/07, James Strachan [EMAIL PROTECTED] wrote:
Sorry for the delay getting back to you. Its taken me a while to get
the time to apply the patch and then get everything
The current subversion head does not support this, however I've
written a patch allowing this. Check AMQ-1241
https://issues.apache.org/activemq/browse/AMQ-1241 for the attached
patch against the current head.
Greetings
Mario
PS: has anybody reviewed the patch?
On 5/15/07, liuxiaoming [EMAIL
and rename that sqlserver file.
Now for the stupid question.
Does that mean I will need to recompile ActiveMQ?
and build a new activemq.jar.
kd
Mario Siegenthaler wrote:
You can tell ActiveMQ to use a specific database type by the following
configuration:
jdbcPersistenceAdapter
You can tell ActiveMQ to use a specific database type by the following
configuration:
jdbcPersistenceAdapter
adapterClass=org.apache.activemq.store.jdbc.adapter.ImageBasedJDBCAdaptor/
Just lookup the correct adapter for SQL-Server. If you get a database
locking issue use the modified adapter by
The default value of networkTTL is 1 (per the documentation); does this
mean
that messages will not cross more than one broker? Could that be causing
the
problem above?
I'd say it's the TTL that causes your problem (it actually does what you
said). Try setting it at least to 2.
I'd also
Newbie question: I just downloaded the binarry install for activemp 4-1-1.
When I go through the 'getting started' steps, I can't access the web
console. the trace doesn't show any reference to a web server (shoud it be
jetty ?)
the http://localhost:8161/admin doesn't seem to be responding.
[.. broker 'losing' every 2nd message.. ]
Oh I just thought of another possible cause (related to the first one): You
might also startup a second broker named localhost (vm://localhost) if you
have renamed you actual broker to something else and use the WebClient.
There's some (IMO weird) code
49 matches
Mail list logo
