Now I need help on how to stop watching this topic (especially since it was
hijacked and is no longer very relevant to my original question, which was
answered satisfactorily six months ago).
The means of removing a watched topic on this forum doesn't appear to work, in
that if I choose "Watche
Hi Justin. I have made another posting to this forum which hopefully explains
my approach: http://www.jboss.org/index.html?module=bb&op=viewtopic&t=63328.
Basically I have written an MBean which uses a TreeCacheAop service to cache
some objects, and it is supposed to synchronously replicate acr
Yes I agree it should be replicating, the problem is that it isn't. I must be
doing something wrong, but for the life of me I can't see what it is. I've
gone through all of the tutorials and reference pages, and my configuration
seems correct. The synchronous replication of the cached objects
In my application I have an MBean which uses the TreeCacheAop service (MBean)
for caching. I am getting synchronous replication only when a new object is
added to the cache. So when I add a new object to the cache it gets replicated
on all servers of the cluster, but when I update member field
The bonehead mistake which was causing the problem: I was always looking up my
EJBs with "java:comp/env/" prefixed to the actual JNDI name of the EJB. Now
that I've removed this everything works as it should.
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewto
A little more information:
The ServiceLocator is a Singleton object which runs as part of the server, and
has its InitalContext bound the first time it's used (this happens in the
ServiceLocator Singleton's constructor). The weirdness I'm experiencing is that
it completes several lookups succes
I have a ServiceLocator class which has a method that looks up and caches EJB
remote home objects. I am occasionally getting a NameNotFoundException from
this method telling me that env is not bound when I try doing a lookup using a
JNDI name such as "java:comp/env/ejb/MyEjb". What is so odd i
The problem was that the class was missing a no-args constructor, and hence
wasn't a valid JavaBean.
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3873463#3873463
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=
I have a JSP which when called gives the following error:
11:27:07,135 ERROR [Engine] StandardWrapperValve[jsp]: Servlet.service() for
servlet jsp threw exception
| org.apache.jasper.JasperException:
/useractivity/UserActivitiesDisplayAndUpdate.jsp(9,0) The value for the useBean
class attribu
I have some code in which I first fetch a cached HashMap from a TreeCacheAop
using getObject(). I then wrap the HashMap as an object of another class which
provides getters and setters to access the HashMap values. Next I add the
object to an HttpSession as an attribute.
The wrapper object's
I had modified the mcast address in the configurations for the two servers on
the first cluster node machine, but only in the configuration of one of the
servers on the second cluster node machine. Bonehead carelessness burned me
good (and for all to see since I posted about it), but fortunatel
Thanks for the responses. A dumb mistake on my part was the problem, and now
that I've corrected it the multiple clusters configuration is working as
advertised.
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3872058#3872058
Reply to the post :
htt
I have two clustered servers which run on the same machines, and I have not
been able to isolate them by modifying the PartitionConfig attribute's UDP
mcast_addr value in the cluster-service.xml. In other words I can have one
application running with nodes on machine A and B, and when I start t
Thanks for your response.
Yes this is exactly what I am doing now, I only wish I had known of this
possibility before wasting so much time trying to use a Singleton. I now have
MBeans which use the TreeCacheAop service with synchronous replication, and
except for a few minor problems it is wor
Why can't someone at JBoss just create the JavaDocs and package them for
download, as is the case for the Java API documentation ?
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3867768#3867768
Reply to the post :
http://www.jboss.org/index.html?modu
Actually on second look that document is almost entirely about HASingletons
running in a cluster. Better to look in the getting started guides or here
http://www.onjava.com/pub/a/onjava/2002/07/10/jboss.html?page=1
The documentation is sparse but in the end basic clustering with JBoss is very
You can look here for a how-to clustering document that I created which is
HASingleton specific. There are all of the basics of clustering included, you
just have to pay attention to what steps are HASingleton specific and skip
those.
http://www.jboss.org/index.html?module=bb&op=viewtopic&t=55
I am getting the below exception when running a clustered application using
JBoss 4.0.1, no idea why. It doesn't appear to be coming from my application
code, although I am calling a QueueSession.commit() when sending as message in
an EJB which might be causing this to happen (can't tell from t
Anil thanks so much for your quick and helpful response. What you suggested
did the trick.
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3866263#3866263
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=3866263
I'm running a JBoss 4.0.1 clustered application and I'm getting the following
console messages every 15 seconds:
| 14:32:37,181 WARN [UDP] discarded message from different group
(Tomcat-Cluster). Sender was brooklyn:32774
| 14:32:37,931 WARN [UDP] discarded message from different group
(
Thanks so much, getting a PropertyResourceBundle as you describe works well.
The initial slash in front of my properties file name didn't help. It seems
that loading Properties isn't as straight-forward as it should/could be with
JBoss. Can it even be done at all ? If so can someone post the
Give some more detailed info and exception messages so we can better help you.
For example do you have the ejb-jar.xml in your JAR ? What are the specific
exceptions ("it blows up" is rather vague) ? etc.
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopi
I have a JBoss application which is deployed in its own directory, for example
JBOSS_HOME/server/myapp/deploy/myapp.sar. In this directory is a
myapp.properties file. I have been unsuccessful at loading the Properties
contained in this file, and I can't figure out why. First I tried a simple
When I try to go to the page with the Oracle AQ as JMS Provider Patch,
http://sourceforge.net/tracker/index.php?func=detail&aid=912173&group_id=22866&atid=376687
I get an error:
ERROR
Artifact: Only Group Members Can View Private ArtifactTypes
Do I need to join a JBoss group to have access t
I have unzipped and untarred the source distribution.
I went into the build directory and ran the build command as suggested:
$ cd jboss-4.0.1-src/build
$ ./build.sh docs
Here is the output and error message:
| Searching for build.xml ...
| Buildfile: c:\Home\Software\JBoss\jboss-4.0.1-sr
I have set up Apache/mod_jk/Tomcat/JBoss in order to route requests to servlets
running on JBoss/Tomcat. It appears that the setup is OK, in that it all works
fine for a VB client (the client's HTTP requests are properly routed to the
servlets with their query strings intact), and it properly r
I have searched for JBoss API documentation but have had no luck. I'm trying
to find JavaDocs such as those available for the J2SE and J2EE APIs. Can
someone tell me whether or not such documentation exists, and if so then what
is the URI to use for downloading it and/or to accessing it online
"monocongo" wrote : Once I finally have everything worked out I will post a forum
message or Wiki entry outlining all of the steps I needed to take to get this working.
Hopefully it will help other developers who want to use Singletons in a clustered
JBoss application avoid the head
Here is an outline of the steps I took in order to use HASingletons in a clustered
JBoss application. I hope it will be helpful to others who try to use HASingletons
for the first time. The below is based on the approach explained by Ivelin Ivanov in
the ONJava article found here:
http://www.
I have created an application which uses "stateful" HASingletons. In the event of the
master node going down these HASingletons are successfully started on the new master
node, but they lose all state information. Once this occurs my application becomes
confused and bad voodoo ensues. I had a
Try this link: http://www.theserverside.com/patterns/thread.tss?thread_id=20819
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3852810#3852810
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=3852810
"lac_raz" wrote : Have tou read the singleton discussion on TheServerSide.com ?
Thanks for making me aware of this discussion. I *really* wish I had seen it before
going down this path with Singletons. My original idea was to use an EntityBean
instead of a Singleton, and from looking at this d
My apologies, I stand rebuked.
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3852409#3852409
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=3852409
---
This
I have just discovered that I'm unable to use three methods which are in the J2EE 1.4
API but which are undefined in the JBoss 3.2.5 library I'm using (jboss-j2ee.jar):
javax.jms.ConnectionFactory.createConnection()
javax.jms.Connection.createSession(boolean, int)
javax.jms.Session.createConsum
This illegal argment error has mysteriously disappeared. The RMIAdaptor now appears
to be working for me as advertised. Hallelujah !
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3852338#3852338
Reply to the post :
http://www.jboss.org/index.html?
I'm using version 3.2.5 on Solaris and Linux machines.
Once I finally have everything worked out I will post a forum message or Wiki entry
outlining all of the steps I needed to take to get this working. Hopefully it will
help other developers who want to use Singletons in a clustered JBoss app
I have worked this out and can access the RMIAdaptor from a non-master node. Here is
the jboss-service.xml of the jmx-invoker-adaptor-server.sar which was copied from
deploy to deploy-hasingleton:
|
|
|
|
|
|
|
|
| jboss:service=invoker,type=jrmp
|
I'm unable to start the NamingService when I set the java.naming.provider.url property
in jndi.properties to the URL for the HA-JNDI tree. When I set this property to
anything other than localhost:1099 (or jnp://localhost:1099, or nothing at all) then
JBoss complains when starting up the Naming
My understanding is that the JBoss HASingleton MBeans do in fact behave as real
singletons (just one instance per JVM) when you run them on a single node of a
cluster. At least they are working well in that capacity for me. I fortunately do
not have the problem of the original poster in that m
"[EMAIL PROTECTED]" wrote : Otherwise its on every node and you have to connect to the
jndi service on the singleton node to lookup the correct RMIAdaptor.
How would I determine at run time which node is the master node ? If I don't know
this then how could I specify a lookup of an object runni
This doesn't work as I was hoping. I have added the below to
JBOSS3.2.5/server/all/deploy/jmx-invoker-adaptor-server.sar/META-INF/jboss-service.xml:
|
|
| jboss:service=DefaultPartition
| jboss.jmx:type=adaptor,name=Invoker
| jboss.jmx:type=adaptor,na
So from what you're saying it appears that the solution is to use the
RMIAdaptor/InvokerAdaptorService as an HASingleton. This way I will always get an
RMIAdaptor which has access to the other HASingleton MBeans as long as I do the lookup
of the RMIAdaptor using HAJNDI in my Servlets and EJBs.
I can't find any reference to a JMX Remote Connector anywhere in the Admin/Devel 3.2.3
Guide, which I assume is the latest Admin/Devel Guide available. Are you referring to
using the InvokerAdaptorService (p. 179) in the same manner as the RMIAdaptor (p. 90)
being used in the RMIAdaptor client
I have just discovered a section in the JMX book (Lindfors, et al) about an RMI
Connector implementation which looks promising. Is this the solution I should be
pursuing ? Any good online references on this topic ?
--James
View the original post :
http://www.jboss.org/index.html?module=bb&o
I have an application which uses several HA Singleton MBeans. It is working well when
running as a single node cluster. But when I run it on additional nodes I can't
locate the singleton MBeans. It seems that the singleton MBeans are only registered
with the MBeanServer on the initial node, a
Here's a good article to get you started:
http://www.onjava.com/pub/a/onjava/2003/08/20/jboss_clustering.html
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3851951#3851951
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mod
This suggestion appears to have gotten me past my original problem, but not out of the
woods. I now get an error with my EJBs not being able to be started in that JBoss is
still looking for DefaultPartition:
| 11:39:19,403 ERROR [StatelessSessionContainer] Starting failed
jboss.j2ee:jndiNa
darrenl perhaps I initially misunderstood your suggestion. Are you suggesting that I
add a PartitionName attribute to the HASingletonControllers, and not just change the
name of the service dependency (depends jboss:service=DevPartition) ?
I will give this a try and see what happens.
--James
I have done this, see my original post above. In fact in every *.xml file which
contained DefaultPartition I have changed the name to DevPartition, including the
jboss-service.xml which is included in my MBeans' SAR. I use the partition name
"DevPartition" in the depends sections for the HASin
anonymous wrote :
| Add a PartitionName attribute to the deploy-hasingleton-service.xml
HASingletonController:
|
As you can see in my original post above I have done this already. Unfortunately it
seems to have no effect.
anonymous wrote :
| Its easier to isolate the cluster traffic
I am attempting to run a cluster using a non-default partition name. In every *.xml
file which originally had "DefaultPartition" I replaced it with "DevPartition". The
jboss-service.xml which is included in the SAR being deployed in the
deploy-hasingleton directory is also using "DevPartition"
I have created a class to send messages to a JBoss queue. It works well but there is
a lot of overhead with each send method call (roughly 60% from what I can tell) since
each time I'm doing JNDI lookups of the QueueConnectionFactory and Queue, and then
establishing the QueueConnection, QueueSe
I'm running 3.2.5 with clustering, and it was relatively painless to move from 3.2.3.
--James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3847705#3847705
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=3847705
--
When I tried adding a queue using an entry like the one specified on page 34 of the
"Getting Started Guide" I ran into trouble, until I figured out that I need to also
have a DestinationManager dependency. So the actual entry needed in
jbossmq-destinations-service.xml should look like this:
I have a MBean which I want to access in other classes. I assume that I should get a
reference to this MBean via a JNDI lookup. If this is true then what name should I
use for the lookup ? Do I need to bind the MBean to a name in the MBean's constructor
or start() method, and then use this na
You're right, I had no understanding of how this actually works under the hood.
You make a very good point that when using the HASingletonDeployer you create a
service and not a true singleton -- I wasn't thinking of it that way. Things are
clearer in that light.
Another point you make which
An off topic reply, but interesting nevertheless.
I think you misunderstand singletons and how they're used, or I do. My understanding
is that a singleton is a class which should never be instantiated more than once in an
application. This is the opposite of what you suggest -- it absolutely d
The documentation I have found regarding the deployment of singletons in a cluster is
contradictory -- in one place it tells you to put your singleton's SAR into the
deploy-hasingleton directory (see
http://www.jboss.org/wiki/Wiki.jsp?page=HASingletonDeployer) but in another place (the
O'Reilly
It appears that it is impossible to deploy a singleton with a private constructor in a
clustered environment. I get the following error when launching the first node of a
cluster in which I have a singleton with a private constructor:
| 15:43:23,917 ERROR [MainDeployer] could not create depl
I am now putting the singleton MBean class and interface in the same
SAR as another MBean I have for my application, instead of in the EAR as
before. The entry for the singleton and controller are now in
JBOSS/server/all/deploy/cluster-service.xml.
When I start JBoss I now get a DeploymentExcept
Am I the only one using singletons with clustering ? It seems to be something which
should have been done many times before, but detailed information on how it works does
not seem to be readily available. I have posted two messages to the JBoss/Clustering
forum (see http://www.jboss.org/index.
Yes you make very good points. When I started to try clustering my servers I thought
running a cluster on one machine would be easier for testing/learning, but now I know
otherwise. Now that I am trying the clustering with two machines there are no more
port issues (though unfortunately there
I am currently trying the approach to Singleton MBeans for clustering suggested in the
JBoss Clustering article by Ivelin Ivanov at
http://www.onjava.com/pub/a/onjava/2003/08/20/jboss_clustering.html, however there is
not enough information in the article -- and I am still working on this approa
I have a message-driven EJB which is listening to a queue which can only have a single
listener (the queue can only be dequeued by one listener) so when I have one node
running in a cluster a second node can't start due to this restriction (at least this
is what appears to be going on). I am ru
Do you need to modify or specify any port numbers for the multiple servers/VMs in a
cluster ? I am assuming that each VM (node) in the cluster will need to listen to
different ports, and this might not be taken care of magically by JBoss, as witnessed
below by my second instance (node) in a clu
I have written an application which I will now need to run in a clustered environment.
I know that now is the wrong time to be considering this, but since the application
is fairly simple I am hoping that it won't need too much modification to enable it to
run in a clustered environment. I am
Thanks for the suggestion.
This was remedied by putting all of the servlets, and not just the first one which did
an initial authentication, into the of the web.xml of the
JBoss web application. The remote user information is now available to all of the
servlets which need it (as long as they
In apache/conf/httpd.conf I am doing a RedirectPermanent to a URL on another machine,
and I think I am losing some session or request information when the request is
forwarded to the redirect URL. I am redirecting to a servlet which is running on a
JBoss server on a remote machine, and in the s
In my MBean class, which extends ServiceMBeanSupport, I do the following in the
startService() method:
| protected void startService ()
| {
| // create a HashTable of environment properties for the InitialContext
| Hashtable environment = new Hashtable();
|
I have tried using several depends in the jboss-service.xml of the SAR file but none
have worked so far - is there a way to do something like depends=all ?
-James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3838400#3838400
Reply to the post :
http://www.j
I have been able to workaround this by changing to the PrefixDeploymentSorter and
prefixing my SAR file name with "1000" -- now the services are sorted with those
starting with digits being deployed later, when 1000UserManager.sar is deployed all
other services have already been deployed.
I mod
Something interesting I have noticed -- the exception doesn't occcur when the SAR file
is not in the deployment directory. I can load the SAR into the deployment directory
after JBoss has already started and everything goes well. So it seems that it is a
problem with the SAR being present when
I'm having an exception come up when I start JBoss, the exception trace is listed
below. This started happening after I copied a SAR file into my deployment directory.
The SAR contains an MBean and a jboss-service.xml. I can't find anything about this
using Google, maybe someone here can shed
This seems to work OK, I can bind the bean to a JNDI name, then access the bean from a
servlet which looks up the MBean using the JNDI name.
-James
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3838384#3838384
Reply to the post :
http://www.jboss.org/inde
I am creating my first MBean and I assume that I need to bind the MBean to a JNDI name
in order to do lookups in my application code to access the MBean. My MBean extends
ServiceMBeanSupport. Should I have some binding code within the startService() method
which will bind the MBean to a JNDI n
I have started getting an exception when I start JBoss, but no idea what is causing
it. Can anyone shed any light on what might be going on (or not) which is causing the
trouble ? Below is the start of the exception listing in the JBoss console (I am
running JBoss 3.2.3 on Windows XP).
I am o
I want to be able to stop and start a message-driven EJB from the JMX console (stop
and start the MDB listening to a queue). Do I need to create an MBean for this ? Any
feedback will be appreciated, thanks in advance.
-James
View the original post :
http://www.jboss.org/index.html?module=bb
The jboss-3.2.dtd gives no applicable information other than that the
destination-jndi-name exists for MDBs, which I was already aware of (see my original
post). What is not clear is how JBoss can make the connection between the JNDI name I
specify in the destination-jndi-name entry and the act
One point I should add is that I am not using JBossMQ for my queue, I am using an
Oracle Advanced Queue.
View the original post :
http://www.jboss.org/index.html?module=bb&op=viewtopic&p=3834518#3834518
Reply to the post :
http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=38345
I'm creating my first Message-Driven EJB and confused as to how the queue that I'll
listen to is specified in JBoss. I have looked at both the "Getting Started" and
"JBoss Administration and Development" guides and neither of these tell you how to
make JBoss aware of an actual queue. For examp
80 matches
Mail list logo
