I have a running JBoss 4.2.3.GA cluster running. My application, not originally
designed for a cluster, has some Singleton objects that hold in memory code
tables. This tables rarely change, but when they do, I need to update the whole
cluster. Since moving to the cluster I face the problem
I have to following code:
| protected TaskInstance locateTask(ContextInstance ctx, JbpmContext
jbpmContext) throws ApplicationException {
| String searchValue = (String) ctx.getVariable(contextKey);
| if (searchValue == null || searchValue.trim().equals())
|
I wish I could recall this issue...
It is my mistake. The environment which had this working had a different
process running which did not run the locateTask code and thus did not fail.
This through me off and caused the question. However, once I found out this
difference I realized that when
I have an existing ejb 2.1 application which I now moved to clustering.
If I understand the documentation correctly 2.1 entities are not shared on the
cluster through a common cache.
This means that if two users running on two different nodes are working on the
same database row through an
I agree both scanarios should be addressed. Addressing the interface issue is
not hard as we can do
type.isInterface()
However this leaves two problems. Both are not common. use cases.
1. Abstract classes which if someone uses it is imposible to get it to work,
since it will fail on
I'm running jbpm 3.1.1 on my site for a long time. Today I tried to see what I
need to do in order to upgrade to 3.2.3. Aside from migrating the DB, I found
several things. One of the things is something I submitted a jira issue
jbpm-696 (which was accepted as a fix and closed).
In
A swimlane, as fas as I understand, will keep the same actor. If TASK B is for
a different actor then a swimlane will not solve this. What we did was to
extend the taskInstance with our own object that keeps the information on the
former actors on the task (Tasks can also be reassigned so you
We are using JBpm 3.1.1 on a big insurance application. We are having memory
problems. After using a profiler we have seen that for some reason all our
actionHandlers(delegations) do not clean from memory. Since these actions are
basically stateless it strikes me as odd.
When a process is
I have an enterprise application running with Jbpm 3.1.1. I have recently done
some profiling work on it and encountered something I can't explain.
Every TaskInsatnce which I get by calling getTaskList on the taskMgmntSession
is never released from memory. So I'm wondering who is holding these
Found my problem. Somehwere along the line I didn't release the context,
everytime I got the new Tasks I created a new context. The former context was
still holding references to the old list and this is why id didn't get cleared
by the GC.
View the original post :
I was able to run with the following configuration:
| # Setup JBoss sepecific properties
| JAVA_OPTS=$JAVA_OPTS -XX:ThreadStackSize=128 -Xmx3072m -Xms1024m
-XX:MaxPermSize=256m
|
| # setup java temporary disk space
| JAVA_OPTS=$JAVA_OPTS -Djava.io.tmpdir=/appdata/jboss/tmp
|
| #
I've been reading through the posts here. I saw that it was recommended in many
places to do the following setting on solaris:
| -XX:ThreadStackSize=128k
|
So I did it! But when I ran it I got an error stating that this new setting of
131072 Kb exceeds the exiting limit of 8192 kB. When I
according to man pages changing the stack limit (which is what I did using
ulimit -s)
is what controlls the number of threads per process. so, I guess my problem
lies elswhere.
View the original post :
http://www.jboss.com/index.html?module=bbop=viewtopicp=4038071#4038071
Reply to the post :
I have a web application running on JBoss 4.0.3SP1 on a sun sparc with 2
processors. Some of my processes are running asynchronously doing alot of
transactions with the DB. When these processes are running the CPU is working
real heavy and the users suffer from a slowdown.
Is there a JBoss
OK. Done alot of investigation but still have a problem. Here is what I got:
first, I saw that my jbmp.configuration has no effect. i.e to have no
transaction support by hibernate i had
| service name='persistence'
factory='org.jbpm.persistence.db.DbPersistenceServiceFactory'
|
I tried as you suggested and commented the flush and release mode and
uncommented the transaction factory. Nothing is changed.
I want to emphasize that everything gets committed or rolledback as it should.
My only problem is that the connection is never closed, thus it is never
returned to
I have recently deployed a large jbpm application. This is actually an
extension to a J2ee application which has been running for two years. Since
this deployment my system exhausted the connection pool every few hours. It
looks like JBPM is responsible (or rather me misusing it)
Anyhow, I
I've created a class that extends TaskInstance. I've created an enhancment
table for it (thought it should be better this way) and done the hbm.xml stuff.
The extension was OK and I was able to create a task. However, I now have these
problems. if I load a task using the JbpmContext (which I
I just checked and I have the exact issue with 1.6.1GA.
I forgot to mention I'm not using the default configuration, I use myu own
copnfiguration. When I start from command line I do:
run -c amit
I chose the 'amit' configuration for the server both in 2.0.0 and 1.6.1 The
server loads, but when
I configured my jboss 4.0.3Sp1 server to run in the ide. This server runs
perfectly from command line. However when I run it from the server and deploy
my application it fails on DB connectivity. My datasource connects using a
login configuration in the conf directory's login-config.xml which
I'm trying to integrate JBpm into an existing j2ee application running on Jboss
in it's own ear. I'm using Jbpm as a service.
I need to write actions that do things directly in the application like
invoking some processes, beans and such. The problem is that Jbpm.sar does not
see my ear file
21 matches
Mail list logo
