2:54 p.m.
[
https://issues.jboss.org/browse/WFLY-9450?page=com.atlassian.jira.plugin....
]
Brian Stansberry edited comment on WFLY-9450 at 10/17/17 3:53 PM:
------------------------------------------------------------------
DeploymentRestResourcesDefintion needs to check that deploymentService is not null.
[~claudio4j] The EJB3 component issue would be a separate issue. TBH I'm not sure
it's really a bug though. HAL should not be trying to recursively read the subsystem
child resources for FAILED deployments. A FAILED deployment is basically in a broken state
so the runtime services underneath it often are not present. The EJB3 resource is
correctly reporting that it can't find the services it needs to respond to the
management request.
Perhaps the kernel could find a way to hide the child resources of a deployment if
it's not DeploymentStatus.OK, but that's not simple to do. It also _might_ break
someone if they had figured out some way to usefully read child resources of a failed
deployment, e.g. to analyze something. I think that's a corner case though.
was (Author: brian.stansberry):
DeploymentRestResourcesDefintion needs to check that deploymentService is not null.
[~claudio4j] The EJB3 component issue would be a separate issue. TBH I'm not sure
it's really a bug though. HAL should not be trying to recursively read the subsystem
child resources for FAILED deployments. A FAILED deployment is basically in a broken state
so the runtime services underneath it often are not present. The EJB3 resource is
correctly reporting that it can't find the services it needs to respond to the
management request.
Perhaps the kernel could find a way to hide the child resources of a deployment if
it's not DeploymentStatus.OK, but that's not simple to do. It also -might- break
someone if they had figured out some way to usefully read child resources of a failed
deployment, e.g. to analyze something. I think that's a corner case though.
NPE on jaxrs subsystem of FAILED deployment
--------------------------------------------
Key: WFLY-9450
URL: https://issues.jboss.org/browse/WFLY-9450
Project: WildFly
Issue Type: Bug
Components: REST
Reporter: Claudio Miranda
Assignee: Alessio Soldano
read deployment resources of a server, where one deployment is in failed status, throws a
NPE and doesn't return the list of deployed applications.
To reproduce, get the kitchensink from quickstarts, change the persistence.xml to point
to ExampleDS datasource, deploy and assign to main-server-group. Then disable ExampleDS,
restart server-one.
Run the following command:
{code}
/host=master/server=server-one:read-children-resources(child-type=deployment,include-runtime,recursive)
{code}
The exception in server.log
{code}
2017-10-17 17:08:02,778 ERROR [org.jboss.as.controller.management-operation]
(ServerService Thread Pool -- 32) WFLYCTL0013: Operation ("read-attribute")
failed - address: ([
("deployment" => "kitchensink.war"),
("subsystem" => "jaxrs"),
("rest-resource" =>
"org.jboss.as.quickstarts.kitchensink.rest.MemberResourceRESTService")
]): java.lang.NullPointerException
at
org.wildfly.extension.undertow.deployment.UndertowDeploymentService.getDeployment(UndertowDeploymentService.java:170)
at
org.jboss.as.jaxrs.DeploymentRestResourcesDefintion$AbstractRestResReadHandler.execute(DeploymentRestResourcesDefintion.java:183)
at
org.jboss.as.controller.operations.global.ReadAttributeHandler.doExecuteInternal(ReadAttributeHandler.java:172)
at
org.jboss.as.controller.operations.global.ReadAttributeHandler.doExecute(ReadAttributeHandler.java:135)
at
org.jboss.as.controller.operations.global.GlobalOperationHandlers$AbstractMultiTargetHandler.execute(GlobalOperationHandlers.java:231)
{code}
The CLI results doesn't display the error message
{code}
[domain@localhost:9990 /]
/host=master/server=server-one:read-children-resources(child-type=deployment,include-runtime,recursive)
{
"outcome" => "success",
"result" => undefined,
"failure-description" => undefined
}
{code}
There is a similar problem while reading a specific deployment,
{code}
/host=master/server=server-one/deployment=kitchensink.war:read-children-resources(child-type=subsystem,recursive)
{code}
results in the following error in server.log
{code}
2017-10-17 17:16:09,857 ERROR [org.jboss.as.controller.management-operation]
(ServerService Thread Pool -- 32) WFLYCTL0013: Operation ("read-attribute")
failed - address: ([
("deployment" => "kitchensink.war"),
("subsystem" => "ejb3"),
("stateless-session-bean" => "MemberRegistration")
]) - failure description: "WFLYEJB0370: EJB component for address [
(\"deployment\" => \"kitchensink.war\"),
(\"subsystem\" => \"ejb3\"),
(\"stateless-session-bean\" => \"MemberRegistration\")
] is in
state DOWN, must be in state UP"
{code}
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)