May 2009 - jboss-dev-forums - Jboss List Archives

[Design of JBoss jBPM] - super-state, scope or something else

by tom.baeyens＠jboss.com

hi guys, i'm starting on the super-state. and i'm in doubt if we should keep the name super-state. the only alternative name i can come up with is scope. i like that a bit better. but we also kept 'state' because it would be easier for people that know jbpm 3, even though 'wait' would have been better. so i think the same applies here and hence we should keep super-state but i just wanted to check if there are other opinions out there. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231244#4231244 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231244

16 years, 10 months

1
0
0 / 0

[Design the new POJO MicroContainer] - Re: Configuration Errors

by trustin

Ah, we are using JAXB. I thought we are something like SAX. Sorry about the noise. Then, what about improving JBoss JAXB itself to maintain a complete map that maps an object property to the location of the XML file? For example: Location l = map.getLocation(object, "bindAddress"); String path = l.getPath(); int lineno = l.getLineNumber(); int column = l.getColumn(); .. This map and location information could be written generic enough so that it can be used with different configuration file formats, as long as the configuration parser (JAXB implementation in the case above) puts the appropriate information when it processes the configuration file. I hope my idea makes sense this time. :) View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231233#4231233 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231233

16 years, 10 months

1
0
0 / 0

[Design of JBoss jBPM] - Re: Timer problem in jBPM

by kukeltje

Wrong forum, use the user forum for this. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231232#4231232 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231232

16 years, 10 months

1
0
0 / 0

[Design of Messaging on JBoss (Messaging/JBoss)] - Re: Potential Netty problem with synchronously closing conne

by trustin

I've just modified Netty so that Channel.close() in channelConnected handler method can close the connection immediately not causing the too many open files error, and started to wonder if you really want to suspend accepting an incoming connection instead of closing them quickly. Even if you suspended an accept operation, the incoming connections are accepted by the operating system and hold in the backlog of the server socket. If the backlog is full, further connection attempt from the client will fail. When you resume the accept operation, the oldest connections will be served while the latest ones have been failed. I'm not sure this is the expected behavior. I would close the connection so that the client can retry later and all connections are served anyway rather than being timed out due to the full backlog. WDYT? View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231220#4231220 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231220

16 years, 10 months

1
0
0 / 0

[Design of JBoss Web Services] - Re: @WebParam sucks, Any workarround or just wait Java7?

by alessio.soldano＠jboss.com

No, there's not such solution in place currently. Feel free to fill in a JBossWS JIRA ticket, perhaps under community contribution component. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231210#4231210 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231210

16 years, 10 months

1
0
0 / 0

[Design of JBoss Web Services] - @WebParam sucks, Any workarround or just wait Java7?

by steeven

To generate a nice WSDL, many stupid @WebParam annotations should be hardcoding to code: public User create( | @WebParam(name = "name") String name, | @WebParam(name = "passowrd" String password); I foud [url=http://paranamer.codehaus.org/]ParaNamer[/ur] has two solution for this issue, the first requires special compiler process, and the second read debug information from class file, see BytecodeReadingParanamer. it should be a fast solution currently. or the 3rd way, post process existing jars + code, add @WebParm with bytecode engineering. Does JBossWS already have such solution? View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231180#4231180 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231180

16 years, 10 months

1
0
0 / 0

[Design of Management Features on JBoss] - ManagedObject persistence for -jboss-beans.xml

by bstansberry＠jboss.com

Emanuel, Still not seeing attachment persistence for the ServiceBindingManager MOs. In Branch_5_x I've modified the way the ServiceBindingManager MOs work along the lines of what we chatted about last week. Basically the conf/bindings-jboss-beans.xml only has one bean that has @ManagedXXX annotations; the rest are pure pojos. That ServiceBindingManagementObject bean exposes a few managed properties, annotated with @MetaMapping. For most properties the result is a CollectionValue whose elements are CompositeValues. This works fine in terms of being able to read the data via the ManagementView and pass back updates to the runtime component. But persistence doesn't work. I get a metadata.attachments.xml like the following but that's it: <?xml version="1.0" encoding="UTF-8" standalone="yes"?> | <attachments-metadata xmlns="urn:org:jboss:profileservice:attachments:1.0"> | <deployment-name>bindings-jboss-beans.xml</deployment-name> | <deployment-structure> | <relative-order>0</relative-order> | </deployment-structure> | <last-modified>1242333438099</last-modified> | </attachments-metadata> I believe the problem is the following call returns null: | Thread [WorkerThread#0[127.0.0.1:41261]] (Suspended) | ManagedDeploymentImpl.getManagedObject(String) line: 212 | AbstractAttachmentStore.updateDeployment(ProfileDeployment, ManagedComponent, boolean) line: 426 | AbstractAttachmentStore.updateDeployment(ProfileDeployment, ManagedComponent) line: 342 | ManagementViewImpl.updateComponent(ManagedComponent) line: 1351 | NativeMethodAccessorImpl.invoke0(Method, Object, Object[]) line: not available [native method] | NativeMethodAccessorImpl.invoke(Object, Object[]) line: 39 | DelegatingMethodAccessorImpl.invoke(Object, Object[]) line: 25 | Method.invoke(Object, Object...) line: 585 | Dispatcher.invoke(Invocation) line: 121 | ProfileServiceInvocationHandler(AOPRemotingInvocationHandler).invoke(InvocationRequest) line: 82 | ProfileServiceInvocationHandler.invoke(InvocationRequest) line: 97 | SocketServerInvoker(ServerInvoker).invoke(InvocationRequest) line: 891 | ServerThread.completeInvocation(SocketWrapper, InputStream, OutputStream, boolean, int) line: 744 | ServerThread.processInvocation(SocketWrapper, InputStream, OutputStream) line: 697 | ServerThread.dorun() line: 524 | ServerThread.run() line: 232 The name of being looked up is "org.jboss.kernel.plugins.deployment.AbstractKernelDeployment". There is no such entry in the ManagedDeploymentImpl.unitMOs map. The component being updated is: org.jboss.profileservice.management.TempManagedComponentImpl@238b1d64{name=ServiceBindingManager, type=ComponentType{type=MCBean, subtype=ServiceBindingManager}, owner=ManagedDeployment@37134967} It's parent is ManagedObject{ServiceBindingManager} whose parent is ManagedObject{org.jboss.kernel.plugins.deployment.AbstractKernelDeployment} View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231178#4231178 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231178

16 years, 10 months

1
0
0 / 0

[Design of JBoss jBPM] - Timer problem in jBPM

by hublisiddu

In our JBPM workflow, from submit request, there will be two node waiting for approval [irrsepctive of priority] once both gets approved then process will be complete. The wait time for approval is 10 minutes, if user doesn't approve in 10 minute, process needs to send email and wait for 10 more minutes, once it reaches total 30 minutes then node should automatically forward to end node. This process will same for both the the nodes. The problem I am facing is, after submitting request when timer starts for both the nodes. in 10 minutes if no one approves the request then timer will check for 30 mins check[whether it finishes 30 mins], sends mail accordingly and either remain in same node or go to end state depending on time check.The samething happens in other node as well. But both try to access the same thread of job executor and creating delegation exception. If I keep different time check for both then it works but if both try to access thread at same time then I am getting delegation exception. If anyone knows possible solution for this problem then their help and guidance is very much appreciated. Thanks in advance, Siddu View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231176#4231176 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231176

16 years, 10 months

1
0
0 / 0

[Design of POJO Server] - Re: MDR AnnotationCreator/AnnotationProxy leads to proxy lea

by alesj

"jason.greene(a)jboss.com" wrote : | Instead we should be passing these routines a short-lived loader to define the proxies in. The AnnotationCreator takes a ctor which has cl as param, we should make sure this gets used. And then the AnnotationProxy should use that one instead of annotation's. I just doubt we always have the right cl present that we could push into AC. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231171#4231171 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231171

16 years, 10 months

1
0
0 / 0

[Design of POJO Server] - MDR AnnotationCreator/AnnotationProxy leads to proxy leakage

by jason.greene＠jboss.com

The current code in AnnotationCreator and AnnotationProxy reuse the classloader that defines an annotation to create the new proxy in. This is bad when the annotation exists in a classloader that stays around for a long time, like say one that would hold jboss annotation classes :) Instead we should be passing these routines a short-lived loader to define the proxies in. View the original post : http://www.jboss.org/index.html?module=bb&op=viewtopic&p=4231169#4231169 Reply to the post : http://www.jboss.org/index.html?module=bb&op=posting&mode=reply&p=4231169

16 years, 10 months

1
0
0 / 0

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

jboss-dev-forums May 2009