[JBPAPP-6505] Removed deployments in farm service prevent node from joining cluster - JBoss Issue Tracker

XML

Word

Printable

Details

Type: Bug
Status: Open (View Workflow)
Priority: Major
Resolution: Unresolved
Affects Version/s: EAP_EWP 5.1.0
Fix Version/s: TBD EAP 5
Component/s: Clustering
Labels:
None

Steps to Reproduce:

Hide

Start one node of a 2-node cluster.
Deploy something to the farm directory, then remove the deployment.
Stop the node.

Start 2nd node of the cluster.
Start 1st node again. It will abort starting with the listed error.

Show
Start one node of a 2-node cluster. Deploy something to the farm directory, then remove the deployment. Stop the node. Start 2nd node of the cluster. Start 1st node again. It will abort starting with the listed error.
Affects:

Release Notes
Workaround:

Workaround Exists
Workaround Description:

Hide

Manually delete the deployment's <content> entry from data/farm-repository-contents.xml before starting the instance.

Show
Manually delete the deployment's <content> entry from data/farm-repository-contents.xml before starting the instance.
Release Notes Text:

Hide
CAUSE: When joining the cluster, the node tries to send all farm deployments it knows about to the other nodes, even ones that have been removed.

CONSEQUENCE: Node cannot join the cluster if it has had a farm deployment removed that the other nodes don't know about.

FIX

RESULT

Show
CAUSE: When joining the cluster, the node tries to send all farm deployments it knows about to the other nodes, even ones that have been removed. CONSEQUENCE: Node cannot join the cluster if it has had a farm deployment removed that the other nodes don't know about. FIX RESULT
Release Notes Docs Status:
Not Yet Documented
Docs QE Status:
NEW

Description

If a node has previously had a farm deployment that is currently removed, that the running cluster members never knew about, it causes the following error on startup when the farm service tries to synchronize with the cluster:

2011-05-11 15:47:47,931 ERROR [org.jboss.system.server.profileservice.repository.ScopedProfileServiceController] (main) Error installing to Create: name=ProfileKey@5ad89e02[domain=default, server=default, name=farm] state=Configured mode=On Demand requiredState=Installed
java.lang.reflect.InvocationTargetException
...
Caused by: java.lang.RuntimeException: java.io.FileNotFoundException: /home/dereed/opt/jboss/instance2/jboss-as/server/all/farm/blah.xml (No such file or directory)
at org.jboss.profileservice.cluster.repository.DefaultRepositoryClusteringHandler.rethrowAsUnchecked(DefaultRepositoryClusteringHandler.java:951)
at org.jboss.profileservice.cluster.repository.DefaultRepositoryClusteringHandler.installModifications(DefaultRepositoryClusteringHandler.java:630)
at org.jboss.profileservice.cluster.repository.DefaultRepositoryClusteringHandler.synchronizeContent(DefaultRepositoryClusteringHandler.java:324)
at org.jboss.system.server.profileservice.repository.clustered.ClusteredDeploymentRepository.load(ClusteredDeploymentRepository.java:255)
at org.jboss.system.server.profile.repository.AbstractProfile.create(AbstractProfile.java:158)
... 22 more
Caused by: java.io.FileNotFoundException: /home/dereed/opt/jboss/instance2/jboss-as/server/all/farm/blah.xml (No such file or directory)
at java.io.FileInputStream.open(Native Method)
at java.io.FileInputStream.<init>(FileInputStream.java:137)
at org.jboss.system.server.profileservice.repository.clustered.local.file.FileReadAction.getInputStream(FileReadAction.java:164)
at org.jboss.system.server.profileservice.repository.clustered.local.file.FileReadAction.getNextBytes(FileReadAction.java:81)
at org.jboss.profileservice.cluster.repository.DefaultRepositoryClusteringHandler.executePush(DefaultRepositoryClusteringHandler.java:785)
at org.jboss.profileservice.cluster.repository.DefaultRepositoryClusteringHandler.executeSynchronizationActions(DefaultRepositoryClusteringHandler.java:673)
at org.jboss.profileservice.cluster.repository.DefaultRepositoryClusteringHandler.installModifications(DefaultRepositoryClusteringHandler.java:592)
... 25 more

It appears the joining node is trying to send the deployment to the other nodes, even though removed=true.

Removed deployments in farm service prevent node from joining cluster

Details

Description

Gliffy Diagrams

Attachments

Activity

People

Dates