Red Hat Training

A Red Hat training course is available for Red Hat JBoss Operations Network

4. Known Issues in JBoss Operations Network

This section contains a subset of the known issues at the time that JBoss Operations Network 3.2 was released. When possible, workarounds are provided.
For a complete list of issues reported with JBoss ON 3.2, check the Bugzilla query for the JBoss ON product, with the version of 3.2.

4.1. Installation and Upgrade

Warning

Upgrade from JBoss Operations Network 3.1 to JBoss Operations Network 3.2 is fully supported.
However, upgrade from JBoss Operations Network 3.2 Beta to JBoss Operations Network 3.2 is not supported. It is not possible to upgrade a beta version of JBoss Operations Network to the release version. If you install a beta server, then you must uninstall the beta server, and then install a server, with all its resource configuration, fresh for JBoss ON 3.2 GA.
Bugzilla #1025782
After upgrading an agent installed by RPM, the rhq-agent_env.sh file is not updated. This means that the rhq-agent_env.sh file — which contains the properties for the agent — contains the path to the old agent installation.
Uninventorying the old agent resource and then re-importing the agent updates the agent information.
Bugzilla #1020896
The following null pointer exception can be logged when the server is installed:
09:07:44,919 ERROR [org.jboss.as.ejb3.invocation] (ServerService Thread Pool -- 67) JBAS014134: EJB Invocation failed on component SchedulerBean for method public abstract void org.quartz.Scheduler.shutdown(boolean) throws org.quartz.SchedulerException: javax.ejb.EJBException: java.lang.NullPointerException
	at org.jboss.as.ejb3.tx.CMTTxInterceptor.handleExceptionInOurTx(CMTTxInterceptor.java:165) [jboss-as-ejb3-7.2.1.Final-redhat-10.jar:7.2.1.Final-redhat-10]
...
Caused by: java.lang.NullPointerException
	at org.rhq.enterprise.server.scheduler.SchedulerService.shutdown(SchedulerService.java:204) [rhq-server.jar:4.9.0.JON320ER3]
Bugzilla #1018238
There is an intermittent problem with installing the server, where the installation returns a JavaException and fails with the error can't connect to 127.0.0.1:6999.
Running the installation process again usually resolves this problem.
Bugzilla #1011254
Depending on the system configuration, installing the server on Solaris 11 fails because the installation process cannot reserve enough space for the object heap.
Resetting the heap size allows the installation process to succeed.
Bugzilla #1027339
The rhqctl status command does not display the correct PIDs for the server and its JVM. It displays the same PID for both processes.
On the system, the server process is named RHQ Server and the JVM process is called JBossAS Java VM child process.

4.2. Windows

Bugzilla #1004900
If a server and storage node are installed on a Windows server, an additional storage node cannot be deployed on a Linux server. The attempted storage node installation fails.
Bugzilla #960166
On Windows, there are limits on how long a pathname can be. When the JBoss ON server archive is extracted, many tools can extract it and create a directory and pathname which is longer than the limit (260 characters). When the pathname is too long, the system cannot locate the files.
Be careful when using the extract all command on Windows. Expanding the archive automatically creates a directory called jon-server-VER.RELEASE/, which is about 20 characters. Using extract all, instead of specifying the directory to which to extract the archive, can double the directory name — for example, C:\example\jon-server-3.2.0\jon-server-3.2.0 — or install it to a downloads directory such as C:\Users\Administrator\Downloads.
It is recommended that you extract the archive to C:\rhq, such as C:\rhq\jon-server-3.2.0.
Bugzilla #1037601
If the JAVA_HOME value contains parentheses, the agent cannot be installed. By default, the JRE is installed in C:\Program Files (x86). An alternative location should be used for installing the agent.
Bugzilla #1037601
If RHQ_AGENT_RUN_AS or RHQ_AGENT_RUN_AS_ME parameter was set in the agent's rhq-agent-env.bat file is set without a password, then the agent upgrade process hangs.
If either of the RHQ_AGENT_RUN_AS* parameters is set, then there must be a password and the password prompt must be disabled.
RHQ_AGENT_PASSWORD=secret
RHQ_AGENT_PASSWORD_PROMPT=false
Alternatively, the RHQ_AGENT_RUN_AS* parameters can be removed prior to upgrading.

4.3. Agents

Bugzilla #1022979
If an agent is installed through an RPM, it cannot be managed using the rhqctl script. The agent RPM is intended only for agents installed on managed platforms. The rhqctl script, on the other hand, is to manage JBoss ON components (the server and storage nodes), so it can only be used to manage agents on a JBoss ON server system.
Bugzilla #1014631
When stopping or starting the agent using rhqctl, the agent log records IllegalStateException errors.

4.4. Server

Bugzilla #1039436
The JBoss ON server cannot be discovered and imported into the inventory. Likewise, attempting to manually import the JBoss ON server as a child of the platform fails.
Bugzilla #1035344
If some resources are being actively imported, then attempting to remove the platform resource or deleting agent fails with the error java.lang.IllegalStateException:Failed to uninventory platform.
Wait until the import operation completes, and then uninventory the platform. If necessary, uninventory all of the child resources for the platform, and then uninventory the platform.
Bugzilla #1029384
A JBoss ON server can be removed from the high availability cloud by removing the server on the Servers page in the UI. The server is successfully removed from the topology, and any agents no longer communicate with that server.
However, if the server is still installed and running on a local system, the web UI for that server can be accessed and it can perform management tasks in the JBoss ON deployment.
To prevent this, uninstall the server from the local system when it is removed from the high availability cloud.
Bugzilla #1026786
When the JBoss ON server is started, it logs a series of JBAS015960 warnings that indicate invalid classpath entries in the deployment. These can be ignored.
Bugzilla #1019456
The server logs record a lot of warnings about not geting consistent results of the paged data:
14:26:52,057 WARN  [org.rhq.enterprise.server.util.CriteriaQueryRunner] (http-/0.0.0.0:7080-6) Could not get consistent results of the paged data and a total count for MeasurementScheduleCriteria[filter
s: [filterResourceGroupId=10021, filterResourceInventoryStatuses=[COMMITTED]], fetches: [], sorts: []]. After 10 attempts, the collection size is 7, while the count query reports 14 for PageControl[page
=0, size=-1, sort[definition.displayName ASC]]. The discrepancy has not cleared up in 3750ms so we're giving up, returning inconsistent results. Note that is most possibly NOT an error. It is likely cau
sed by concurrent database activity that changes the contents of the database that the criteria query is querying.: java.lang.Exception
at org.rhq.enterprise.server.util.CriteriaQueryRunner.execute(CriteriaQueryRunner.java:81) [rhq-server.jar:4.9.0.JON320ER3]
Bugzilla #1026799
If LDAP authentication is enabled, there are spurious warnings about invalid modules for each LDAP configuration property.
09:17:31,266 WARN  [org.jboss.security] (http-/0.0.0.0:7080-5) PBOX000234: Invalid or misspelled module option: BindDN
09:17:31,266 WARN  [org.jboss.security] (http-/0.0.0.0:7080-5) PBOX000234: Invalid or misspelled module option: Filter
09:17:31,266 WARN  [org.jboss.security] (http-/0.0.0.0:7080-5) PBOX000234: Invalid or misspelled module option: java.naming.factory.initial
09:17:31,266 WARN  [org.jboss.security] (http-/0.0.0.0:7080-5) PBOX000234: Invalid or misspelled module option: LoginProperty
09:17:31,266 WARN  [org.jboss.security] (http-/0.0.0.0:7080-5) PBOX000234: Invalid or misspelled module option: BaseDN
09:17:31,266 WARN  [org.jboss.security] (http-/0.0.0.0:7080-5) PBOX000234: Invalid or misspelled module option: GroupFilter
09:17:31,267 WARN  [org.jboss.security] (http-/0.0.0.0:7080-5) PBOX000234: Invalid or misspelled module option: java.naming.provider.url
Bugzilla #1026836
If the backend database is stopped while a scheduled quartz job (such as a bundle deployment) is running, the quartz job cannot be rescheduled and never completes, even after the backend database is restarted and running.

4.5. Platform Plug-in

Bugzilla #1026513
File system child resources for a platform are shown as available even if they are offline. This can occur if a file system directory is unmounted or removed after being imported.

4.6. EAP 6 Plug-in

Bugzilla #1003114
EAP 6.0.0 and EAP 6.1.0 handle the keep alive value for the thread pools differently. On EAP 6.1.0, the unit is case insensitive, so the value accepts both seconds and SECONDS. On EAP 6.0.0, only all-capital letters are accepted (SECONDS). If a lower-case string is used, it returns an operations error, Operation handler failed: No enum constant.
Bugzilla #1026437
A datasource must be disabled before its configuration can be edited in the JBoss ON UI. However, the server incorrectly returns an error that a datasource is active and must be disabled when attempting to change the configuration, even if the datasource has been disabled. The configuration update is rejected.
Bugzilla #1012057
The UI allows a user to edit the resource configuration of a deployment scanner to point to a different directory name. However, if that directory does not already exist on the system, the configuration update fails (even though it appeared to succeed).
Bugzilla #1016678
If an agent discovers an EAP 6 server but the HTTP management interface definition is not included in its configuration file (such a host-slave.xmlM), then the agent incorrectly sets the management port to 0. Since the management port is not actually 0, the agent cannot connect to the EAP server.
Bugzilla #1030280
For the EAP 6 server platform MBean, the memory pool child resource metrics charts have the Y-axis improperly labeled as 00.000 rather than the real values.
Bugzilla #1033404
Updating the blocking bounded queue or bounded queue thread pool configuration properties forces a change to the value for core threads for the resource. Otherwise, the configuration update attempt fails.
Bugzilla #1014668
The XADataSource resource has an empty XA Datasource class value in the configuration page. When editing or viewing the resource configuration, it forces the user to set a value for the class or navigating away from the pages fails with the error The following configuration properties have invalid values: [Xa Datasource Class]. The values must be corrected before the configuration can be saved.
Bugzilla #1016698
The reload operation for the EAP 6 server fails.
Bugzilla #1013713
Whne events monitoring is enabled with a date format other than the default, the EAP 6 server becomes unavailable.

4.7. EAP 5 Plug-in

Bugzilla #1025812
Changing any value in the ExpiryQueue configuration for the JBoss Messaging resource requires a value for the expiryQueue field to be set explicitly as well, or saving the configuration fails with an exception:
Failed with exception: java.lang.reflect.UndeclaredThrowableException:Failed to set property 'expiryQueue' on component 'jboss.messaging.destination:name=DLQ,service=Queue' to value [SimpleMetaType:java.lang.String:null]. -> javax.management.RuntimeOperationsException:null -> java.lang.IllegalArgumentException:null object name
Bugzilla #1015501
Changes made to JMS queue and topic configurations are not persisted after the resources are restarted. A class cast exception is recorded in the server logs:
2013-10-17 06:33:06,299 WARN  [org.jboss.system.server.profileservice.persistence.deployer.ProfileServicePersistenceDeployer] (main) Failed to update the persisted attachment information
java.lang.ClassCastException: org.jboss.system.server.profileservice.persistence.xml.PersistedSimpleValue cannot be cast to org.jboss.system.server.profileservice.persistence.xml.PersistedCompositeValue
	at org.jboss.system.server.profileservice.persistence.AbstractValueRecreation.createMetaValue(AbstractValueRecreation.java:187)
Bugzilla #1019262
A datasource, queue, or topic child resource is not removed from an EAP 5 server until after the server is restarted.

4.8. EAP 4 Plug-in

Bugzilla #999494
When manually importing an EAP or EAP 5 resource into the inventory, the job is submitted and appears to complete successfully, but the EAP servers are not added to the inventory.

4.9. UI

Bugzilla #1020874
The Add subscriptions button the Content->Subscriptions tab for a platform resource returns an HTTP Status 404 error if the JBoss patch feed is selected. There are no errors recorded in the server log.
Bugzilla #959455
The number of rows shown on the Administration->Agents page are not lazily loaded. This means that it could display a much lower number when the page is initially loaded, and then the number increases as the user scrolls past the loaded rows.
Bugzilla #952665
When creating a new local disk storage content source in the Administration->Content area, click the local disk storage link returns an IllegalStateException.
Bugzilla #1029389
If a server is selected to be removed and the removal is not confirmed (the No is clicked), then all of the buttons on the Administration->Servers page are disabled.
Bugzilla #1020884
When uploading new content to a platform resource, clicking the Continue button after uploading the package returns the error The page requested cannot be displayed due to some error.

4.10. Provisioning

Bugzilla #1026473
Normally, the compliance=full setting in a bundle deploy.xml file instructs the provisioning process to remove any existing files and directories from the target directory and then copy in the bundle files. However, any existing empty directories are not removed when the bundle is provisioned.
Bugzilla #1015406
In some deployment scenarios, attempting to revert a bundle deployment results in a NullPointerException. Additionally, the revert wizard itself becomes unusable.
Bugzilla #1038597
Occasionally, there can be problems with database locks and the quartz job scheduler which prevent a bundle from being deployed successfully. These errors are recorded in the agent log.
Bugzilla #1029121
If a bundle group is assigned to a role, it cannot be deleted. Attempting to delete it fails with a ConstraintViolationException.
First remove the bundle group from any roles, and then delete it.

4.11. Monitoring

Bugzilla #1028623
When measurement data are sent to the server, the alert condition processing relies on the order to be strictly the same as generated on the agent side. Meaning, to assess the condition, no datapoint should be processed before an older (preceding) datapoint.
However, the persistence in the metrics storage node is asynchronous. This means that datapoints can be persisted in an order that is different than the order they were sent by the agent.
When evaluating data to process alert conditions, the datapoints are evaluated in the order in which they are persisted in the storage node, not the order in which they were created on the agent.
It is possible that this could lead to false positives to trigger an alert or to skew dampening events.
Bugzilla #1028944
For an autogroup Monitoring tab, the values displayed on the metrics tables for minimum, maximum, and average readings do not match the values charted or displayed on the individual graphs.
Bugzilla #1031820
In previous versions of JBoss ON, it was possible to override the calculated baselines for a monitoring graph and set a manual baseline. However, that ability was removed from the monitoring graphs in JBoss ON 3.2.
Bugzilla #1035890
If an alert definition is set for a trait on a resource, and the value of that trait is modified using the JBoss ON REST API, no alert is triggered.
Bugzilla #1033347
When setting a custom date range for a graph in the Monitoring->Metrics tab of a resource, the label for the graph displayed is for the previous date range, not custom date range.
However, the data displayed in the graph is properly updated to the new range.
Bugzilla #1034878
When using a custom data range, there is an intermittent error on the resource Summary->Activity page which prevents the monitoring graphs from being loaded.