[jboss-jira] [JBoss JIRA] (WFLY-975) Deployment replace causes invalid state in DUP

Deployment replace causes invalid state in DUP ---------------------------------------------- Key: WFLY-975 URL: https://issues.jboss.org/browse/WFLY-975 Project: WildFly Issue Type: Bug Security Level: Public(Everyone can see) Components: Server Reporter: Thomas Diesler Assignee: Stuart Douglas Fix For: 8.0.0.Alpha1 Attachments: v200.jar, v201.jar, webapp-v200.war Webapp A depends on deployment B. Replacing B causes NPE in JPAInterceptorProcessor {code} 12:52:05,712 INFO [org.jboss.as.server.deployment] (MSC service thread 1-1) JBAS015876: Starting deployment of "v200.jar" (runtime-name: "v200.jar") 12:52:05,889 INFO [org.jboss.as.server] (XNIO-1 task-8) JBAS018559: Deployed "v200.jar" (runtime-name : "v200.jar") 12:52:37,207 INFO [org.jboss.as.server.deployment] (MSC service thread 1-5) JBAS015876: Starting deployment of "webapp-v200.war" (runtime-name: "webapp-v200.war") 12:52:37,349 INFO [org.jboss.web] (ServerService Thread Pool -- 52) JBAS018210: Register web context: /webapp-v200 12:52:37,436 INFO [org.jboss.as.server] (XNIO-1 task-3) JBAS018559: Deployed "webapp-v200.war" (runtime-name : "webapp-v200.war") 12:53:20,252 INFO [org.jboss.web] (ServerService Thread Pool -- 57) JBAS018224: Unregister web context: /webapp-v200 12:53:20,272 INFO [org.jboss.as.server.deployment] (MSC service thread 1-1) JBAS015877: Stopped deployment v200.jar (runtime-name: v200.jar) in 25ms 12:53:20,274 INFO [org.jboss.as.server.deployment] (MSC service thread 1-5) JBAS015876: Starting deployment of "v200.jar" (runtime-name: "v200.jar") 12:53:20,293 ERROR [org.jboss.msc.service.fail] (MSC service thread 1-2) MSC000001: Failed to start service jboss.deployment.unit."webapp-v200.war".FIRST_MODULE_USE: org.jboss.msc.service.StartException in service jboss.deployment.unit."webapp-v200.war".FIRST_MODULE_USE: JBAS018733: Failed to process phase FIRST_MODULE_USE of deployment "webapp-v200.war" at org.jboss.as.server.deployment.DeploymentUnitPhaseService.start(DeploymentUnitPhaseService.java:127) [jboss-as-server-8.0.0.Alpha1-SNAPSHOT.jar:8.0.0.Alpha1-SNAPSHOT] at org.jboss.msc.service.ServiceControllerImpl$StartTask.startService(ServiceControllerImpl.java:1972) [jboss-msc-1.1.1.Final.jar:1.1.1.Final] at org.jboss.msc.service.ServiceControllerImpl$StartTask.run(ServiceControllerImpl.java:1905) [jboss-msc-1.1.1.Final.jar:1.1.1.Final] at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145) [rt.jar:1.7.0_13] at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615) [rt.jar:1.7.0_13] at java.lang.Thread.run(Thread.java:722) [rt.jar:1.7.0_13] Caused by: java.lang.NullPointerException at org.jboss.as.jpa.processor.JPAInterceptorProcessor.deploy(JPAInterceptorProcessor.java:52) at org.jboss.as.server.deployment.DeploymentUnitPhaseService.start(DeploymentUnitPhaseService.java:120) [jboss-as-server-8.0.0.Alpha1-SNAPSHOT.jar:8.0.0.Alpha1-SNAPSHOT] ... 5 more {code} What might be happening here is that replacing B causes ModuleB to go down. Phase FIRST_MODULE_USE of A depends on ModuleB, which causes a call to undeploy() on the DUPs for A. When ModuleB becomes available again, deploy() is called on the DUPs for A again. Possibly because of the cleanup phase, necessary state is lost and we see a NPE. AFAICS, this is a variation of the feature request that DUPs support multiple calls to deploy/undeploy.