I must correct my previous information - on JBossAS 5.0.1 it's working correctly, so it's problem with JBossAS 5.1 only.
I had similar problem on my notebook. It was probably caused by turned off swap on my ubuntu and finally out of memory.
I added memory for JBoss (... -Xms512m -Xmx1024m) and it unfortunately changed nothing. Strange is that few first deployments the project worked correctly (without redeployments) and it changed when I added url rewrite to it. Further changes (i.e. removing url rewrite) didn't help.
I tried to deploy my application (generated by seam-gen from Seam 2.1.2.GA) on JBossAS 5.1.0.GA with hope of no infinite redeployment loop. Unfortunately it's behaving the same way as on JBossAS 5.1.0.CR2.
Does somebody have the same problem as mine?
I did see this the other day in a Seam app that I had generated for a test. I'm running Seam 2.1.1, Eclipse 3.4, JBoss 5.1.0, JDK 6.0. I would deploy the app and it would continually redeploy, rendering it unusable. The project was generated using JBoss Tools in eclipse (which uses seam-gen internally). I wasn't able to determine exactly what was causing it and had to recreate the project. Sorry I don't have a solution for you, but I do feel your pain. Things I tried:
- Deleting jboss/server/all/data, tmp, etc.
- Closing / Opening project in Eclipse, restarting eclipse
- Clean all (obviously)
- Deleting deployed app and doing a full redploy
One thing I didn't try as a test was to disable the HDScanner hot deployer in JBoss to see if that had something to do with it. I didn't have anything
touchingmy descriptors that I know of. Also be aware that in JBoss 5.x there is a seam.deployer in jboss/server/all/deployers that I believe does some hot deployment. I've yet to come across an explanation of what exactly it's doing though.
I should also note that the app/project was working 100% correct for a few days, then this occurred completely out of the blue. If you figure it out, please let me/us know.
I just happened to come across this, you might want to jump on this discussion as well.
Richard, thank you for your suggestions.
I started the EAR undeploys in JBAS 5.1.0GA thread and worked with some of the jboss guys to figure out what was going on.
They opened JIRA-192 which has been closed. I followed the guys instructions and did the following to updated the deployer
svn co http://anonsvn.jboss.org/repos/jbossas/projects/jboss-deployers/trunk/ mvn clean install The trunk/deployers-vfs-spi/target folder will then contain jboss-deployers-vfs-spi.jar which you can use to overwrite the one in JBOSS_HOME/lib folder.
But the issue is still occurring. I emailed him back but haven't heard back yet...only been a few hours though.
How does this happen in a GA release? This seems like a really bad and basic (easily noticeable you'd think!) problem...
I tried to remove SeamMTMatcher bean from seam-deployers-jboss-beans.xml (JBossAS 5.1.0.GA) as it's been described in this topic: http://www.seamframework.org/Community/JBossSeamAndJBossAS5 and, at least for now, it solved problem of never ending redeployment.
I don't know what side effects should I expect due to this removal.
I'm experiencing the exact same issue. Non-exploded deployment works just fine. Seems very odd.
FTR -- I just encountered this. Removing SeamMTMatcher worked for me, too.
For what it's worth, the following post solved this problem for me:
JBoss User Forum
I also had .jsfdia and .spdia files in my exploded war's WEB-INF directory, which seemed to cause JBoss to infinitely redeploy the war.
Excluding those files (apparently they are created when editing pages.xml or faces-config.xml with JBoss Tools) when exploding the war let JBoss 5.1.0.GA deploy the exploded war correctly.