IBM Support

Release notes for IBM PureApplication System W1500 and W1700 Version 1.1.0.1

Release Notes


Abstract

This release notes document contains information about known issues in this release.

Content

After updating systems to Version 1.1.0.1 and applying the Version 1.1.0.1 Group Fix, you can update the System Monitoring shared service on existing deployments. However, do not update the Database Performance Monitor shared service or the IBM Endpoint Manager shared service.

The following issues are known in this release.

Administering the system

Command-line tool

Disaster recovery

Backup and restore

Managing virtual systems and virtual applications

IBM Image Construction and Composition Tool

Database performance monitoring

Advanced Middleware Configuration




Administering the system



Problems occur when applying WebSphere Application Server interim fix to an upgraded virtual system instance
ProblemWhen applying a WebSphere Application Server interim fix to an upgraded virtual system instance, the service history might show the following failed status: Unable to apply service on virtual machine. 
ResolutionFollow the procedures for downloading and installing interim fixes in the WebSphere Application Server product documentation. Interim fixes for your version of WebSphere Application Server are available for download on IBM Fix Central at http://www.ibm.com/support/fixcentral/. For installation procedures, see the WebSphere Application Server Information Center for your product version, for example:

New deployments do not appear immediately on the product license page
ProblemNew deployments do not immediately appear in the In the cloud now section on the product license page. New deployments are recognized by the license timer every 30 minutes to ensure that no erroneous entries exist.
ResolutionWait 30 minutes, and check the product license page again.

Missing 'DisplayItem' choices in monitoring agent for PureApplication System situation definitions
ProblemWhen creating situation definitions by using data from the monitoring agent for PureApplication System, the following error message displays when you attempt to specify a display item in the situation editor:

KFWITM047E. The display item feature can not be used with this situation because no attributes are eligible for use as a  display item in group

The DisplayItem field in a situation definition allows you to identify an attribute that uniquely identifies event instances that are associated with various resources that are being monitored by the situation.  

Although some of the attribute groups in this agent have DisplayItem choices, other groups have only few or no DisplayItem choices.  
ResolutionThis is a known limitation in this release.

PureSystems Manager seems to reboot after being powered off 
ProblemWhen using the command-line tool or physical power button on the hardware to power down the active PureSystems Manager, the PureSystems Manager turns back on after a brief delay of five to 30 minutes. The active PureSystems Manager powers down but is automatically restarted by the other PureSystems Manager.

Due to the power-off action on the active PureSystems Manager, a failover occurs and processes are started on the non-active PureSystems Manager. When one PureSystems Manager starts when the other one is powered down, the first PureSystems Manager automatically starts the second one.  

When powering down the non-active PureSystems Manager, it does not reboot because a failover does not occur. 
ResolutionPower down the PureSystems Manager in the system console. Click Hardware > Management Nodes, select the desired PureSystem Manager, and click Power Off.

Powering down a PureSystem Manager from the system console does not cause it to reboot, regardless of whether it is the active or non-active node.

Exception occurs when adding 90 or more VLANs to the customer network configuration
ProblemPureApplication System automatically defines a maximum of 127 spanning tree group numbers for each VLAN that is assign to a spanning tree group. When the total number of VLANs exceed 127, the spanning tree is automatically disabled.  In this design, PureApplication System does not support more than 90 VLANs in the VLAN range.

To manage virtual local area network (VLAN) information, open the system console and click Hardware > Customer Network Configuration. In the VLAN Information section, you can specify VLAN ranges and enable spanning tree (PVRST). 
ResolutionDo not define more than 90 VLANs when specifying the VLAN range.

The 'Physical cores' view of a compute node incorrectly shows 32 cores per CPU
ProblemFor PureApplication System W1700 systems, the Physical cores view for compute nodes incorrectly shows 32 cores per CPU. However, the correct value is eight cores per CPU.  

To see the physical cores for a compute node, perform the following steps:

1. Open the system console.
2. Click Hardware > Compute Nodes.
3. Select a compute node in the left navigation pane.
4. In the main window, expand the Physical cores section.
ResolutionThe incorrect value of 32 does not affect your deployments. The value only displays incorrectly in the user interface and in the REST API informational messages that describe the physical characteristics of CPU modules.

External monitoring agents for Monitoring Service shared service do not display correctly
ProblemThe monitoring agents for WebSphere Application Server, HTTP Servers, MQ Server, and Message Broker do not support the Self Describing Agent (SDA) feature of IBM Tivoli Monitoring (ITM) 6.2.3 and later when using the external System Monitoring shared service in PureApplication System 1.1.0.1, 1.1.0.0, and 1.0.0.x.  If you have not manually installed the application support for the monitoring agents on the external ITM Hub and Tivoli Enterprise Portal (TEP) servers, then the monitoring agents do not display the proper messages from the workspaces/views/situations/messages directory on the TEP console. 

Also, if you have deployed any existing workloads prior to PureApplication System 1.1.0.0 or any workloads that use versions of the Foundation pattern prior to 2.0.1.0, those workloads might not display correctly if the application support for the monitoring agent for Workloads or the monitoring agent for Linux OS and Unix OS version 6.2.2.7 or later is not installed.
ResolutionTo correct this problem, manually install the application support for each of the monitoring agents, OS agents, and monitoring agent for workloads that is used on the external ITM Hub and TEP servers.  Additionally, install the application support for the monitoring agent for WebSphere Application Server on all Remote-TEMS servers in the external shared service.

For more information, call IBM Support.

System Monitoring service for WebSphere MQ Server is not supported on PureApplication System W1700
ProblemPureApplication System W1700 does not support the System Monitoring service for WebSphere MQ Server.
ResolutionThis is a known restriction.

System Monitoring service for WebSphere Message Broker is not supported on PureApplication System W1700
ProblemPureApplication System W1700 does not support the System Monitoring service for WebSphere Message Broker.
ResolutionThis is a known restriction.

PureSystems Manager might reject upstream Network Time Protocol (NTP) servers running on Windows operating system
ProblemIf PureApplication System is configured to synchronize its time with Network Time Protocol (NTP) servers that are running on the Windows operating system, the PureSystems Manager might reject these servers due to the failure of the distance criteria. The error is 'flash=400 peer_dist', which is seen when using ntpq with pe, as, and rv <association_id> commands.
ResolutionMaintain an accurate clock in PureApplication System. Since the W32Time service is not a full-featured NTP solution that meets time-sensitive application needs, synchronize the PureApplication System with NTP servers that are running on the Linux/Unix operating system.

For more information about this issue, see the Microsoft Support site.

WASError message displays in Monitoring Portal on newly created WebSphere Cluster
ProblemAfter adding a WebSphere Cluster that is connected to the System Monitoring Portal, the Monitoring Portal indicates that an error exists in the WebSphere log file.   
ResolutionAfter reviewing the error message in the WebSphere log file and taking the required actions, ignore the message in the System Monitoring Portal.

Manually powering off a compute node causes an automatic restart
ProblemFor PureApplication System W1700 systems, if you press the Power Off button on the hardware, the system reacts to this unexpected event by restarting the compute node. This automatic restart occurs because the system expects compute nodes to be powered off through the user interface, not by pressing the physical Power Off button on the hardware.
ResolutionDo the following steps in the system console to power off the compute node:
1. Click Hardware > Compute Nodes.
2. Locate the compute node in the left navigation pane.
3. Click Power Off to start the quiesce and power off the compute node.

Unexpected error occurs after deploying a WebSphere Application Server cluster on a system with monitoring enabled
ProblemAfter deploying a WebSphere Application Server cluster on a system with the System Monitoring shared service running, you might see an error similar to the following one in the System Monitoring Portal:

KQ9_Availability_Fail_Sit

Situation Description:  Executes when the role is in a failed availability state.
ResolutionThis is a known limitation.

Deployments fail with DNZVMP481E error message on PureApplication System W1700
ProblemWhen using PureApplication System W1700, deployments might fail with the following error message:

DNZVMP481E Unable to configure new physical volume devices on VIOS Server for the newly allocated SAN LUNs. This may be caused by pre-existing physical volumes having the same initiator WWPNs, LUN addresses, and target WWPNs as the newly allocated volumes.
ResolutionCall IBM Support to resolve the issue.

PureApplication System base operating system image uses IPv6 to connect to customer data networks
ProblemThe base operating system image that is provided by default with PureApplication System W1700 uses Internet Protocol version 6 (IPv6) to connect to customer data networks. However, certain enterprises might not want their data networks connecting through IPv6 for security reasons.
ResolutionDo the following steps to extend and capture the base operating system image so that it does not use IPv6 to connect to data networks:

1. Run this command to connect to your virtual machine and log in as root: ssh root@<ip_addr>.
2. Run this command to open the rc.tcpip file in edit mode: vi /etc/rc.tcpip.
3. Locate the following section in the rc.tcpip file:

# Start up autoconf6 process
start /usr/sbin/autoconf6 "" " -i en1 "
start /usr/sbin/autoconf6 "" " -i en0 "
#start /usr/sbin/autoconf6 "" 

4. Comment out this line: start /usr/sbin/autoconf6 "" " -i en1 "
5. Save your changes.
6. Capture the image with your changes and use it in your deployments. 

HADR feature in IBM DB2 for BLU Acceleration Pattern type is not supported 
ProblemThe Web Application Pattern 2.0.1.0 type that is shipped with PureApplication System 1.1.0.0 and 1.1.0.1 does not support the high availability disaster recovery (HADR) feature that is included in IBM DB2 for BLU Acceleration Pattern 1.1.0.8 type, which is downloaded separately from PureApplication System.
ResolutionThis is a known restriction.


Command-line tool



Extra default configure NIC add-ons are imported into a pattern after running the importPatterns.py script
ProblemAfter creating a virtual system pattern from an image with multiple network interface controller (NIC) cards and then exporting the pattern from PureApplication System by using the exportPattern.py script, the parts section of the patterns.json file might contain extra entries in the add_ons sections for the corresponding parts, for example:

"add_ons": [
               {
                 "add_on_ref": 0,
                  "parameters": [
                  ]
               }
             ]

Then, if you import this same virtual system pattern into PureApplication System with the importPatterns.py script, the pattern contains extra default configure NIC add-ons.
ResolutionRemove the extra add-on entries from the patterns.json file and run the importPatterns.py script again. The patterns.json file is located in the directory or TGZ file that you specified when exporting the virtual system pattern.


Error occurs when running the startup command for the command-line tool
ProblemWhen using a Microsoft Windows operating system, after downloading the PureApplication System command-line tool and extracting the contents of the ZIP file, the following error might occur when you run the startup command:

Cannot find or load org.python.util.jython

This error occurs because the target directory for the extraction contains the term bin in the path name, for example C:\\tools\bin\,
ResolutionDo not extract the command-line tool ZIP file into a directory that includes the term bin in the path name.

Disaster recovery



Simplified disaster recovery disk replication provides application consistency only within cloud groups
ProblemThe application consistency provided by disaster recovery disk replication provides point-in-time consistency only when recovering all of the virtual machines within each cloud group. After all storage volumes for a cloud group have completed initial copying and reached an available state, all disk writes within that cloud group are replicated in consistent order from the primary system to the backup system.

If your deployment spans more than one cloud group, or your application contains tiers that span more than one cloud group, disaster recovery disk replication does not provide point-in-time consistency across your entire deployment or application.
ResolutionIf your application supports inconsistent ordering of disk writes between its virtual machines or tiers, no changes are required.

If your application requires point-in-time data consistency across all virtual machines and tiers, reconfigure your workload so that all virtual machines and tiers run within the same cloud group. This reconfiguration requires you to delete portions of your deployment or application and redeploy them within another cloud group.


Backup and restore



Current limitations when using REST API to obtain estimated size of system backup
ProblemTo retrieve the estimated space requirement of a system backup, you can log in to PureApplication System as the default administrator user and use the following REST API call:

https://<RACK_IP>/admin/resources/backup/backupsize 

<RACK_IP> is the IP address of the system that is being backed up.

Note the following limitations when using this REST API call:
  • The estimated size does not include the size of other components that are exported, for example virtual images and virtual appliances.
  • The estimated size does not differentiate between a base and a differential backup size.
ResolutionDo the following actions to work around these limitations: 
  • If exporting default images that are provided with the system, add an additional 50G to the backup size estimate. 
  • If exporting custom virtual images or virtual appliances, take into account the size of those components when exporting with the backup. For example, when exporting custom images with attached disks and volumes, look at the size of the attached disk and volume and add that number to the backup size estimate.
  • For differential backups, be aware that the backup size might be smaller when running a subsequent backup to the same remote host location and path.

New certificate-key pairs cannot be used with existing system backups
ProblemAfter generating a certificate-key pair and performing a system backup, if you regenerate the certificate-key pair and then attempt to perform another backup, the following exception might occur:

java.lang.Exception: Status 400 returned by POST on localhost/resources/backup during workload backup, output: Unable to create a differential backup where there exists a baseline backup encrypted with different keys
ResolutionOnly regenerate a new certificate-key pair when changing the backup storage path and creating a baseline backup.

Managing virtual systems and virtual applications



Deployment failures occur with WebSphere Application Server after installing System Monitoring Pattern Type 1.0.1.1
ProblemAfter installing the System Monitoring Pattern Type 1.0.1.1 that is available with PureApplication System 1.1.0.1, virtual application or virtual system deployments of WebSphere Application Server fail for cloud groups that have the System Monitoring for WebSphere Application Server shared service deployed, even if the System Monitoring shared service is not upgraded to the new level.  This failure is due to script updates that are installed on new deployments that expect a newer level of the monitoring agent for WebSphere Application Server data collector; however, the System Monitoring for WebSphere Application Server shared service still references the prior agent binaries.
ResolutionPerform the following steps to resolve the failures for new deployments:

1. Delete the System Monitoring for WebSphere Application Server shared service. This deletion causes existing WebSphere Application Server instances that are being monitored to restart after the monitoring agent is uninstalled.
2. Wait for all affected deployments to return to 'Running' state. Wait time is typically five to 15 minutes, depending on the application pattern. 
3. Redeploy the System Monitoring for WebSphere Application Server shared service. Redeploying the shared service causes existing WebSphere Application Server instances to restart after the monitoring agent is reinstalled.

Delay occurs when migrating virtual machines after a compute node is physically removed  
ProblemIf a compute node is mistakenly removed from the mini version of PureApplication System W1700, you might notice that some virtual machines remain in stopped state for 30 minutes or longer before they are migrated to another compute node in the same cloud group. 
ResolutionThis is a known limitation.

Brief throughput failures after applying an emergency fix to a virtual system with an ODR dynamic cluster and APC elasticity enabled
ProblemIf a deployed virtual system is using on demand router (ODR) dynamic cluster and application placement controller (APC) elasticity, and if an emergency fix is applied after the initial deployment, the elasticity addition to the ODR dynamic cluster might trigger throughput errors.
ResolutionTo apply required emergency fixes to extended images and avoid throughput errors, see Applying required fixes for virtual images in Version 3.1.0.2 and complete steps 1-6.

Increasing allocated CPU or memory resources might require stopping or rebooting the database or virtual machine
ProblemWhen using IBM PureApplication System W1500, certain limitations exist when scaling the allocated memory or number of CPUs of a deployed virtual machine that is powered on. You might see the error CWZIP1987E in log files.

Because of these limitations, you might need to take additional steps when any of the following situations apply:

* The allocated memory is less than or equal to 3 GB and you want to increase it to a value greater than 3 GB
* You want to increase the allocated memory to more than 16 times the memory allocation when the virtual machine was last started
* The current CPU allocation is one virtual CPU and you want to add more virtual CPUs
* You want to increase the CPU allocation to a value greater than eight virtual CPUs
ResolutionPerform the following steps to modify the allocated virtual CPUs or memory size of the virtual machine:
1. Stop the virtual machine instance.
2. Resize the allocated memory or number of virtual CPUs.
3. Restart the virtual machine instance.

Memory utilization values differ between the 'Virtual Machine' and 'Virtual Application' pages
ProblemWhen viewing memory utilization values on the Virtual Machine page or Virtual Machines section of the pages for virtual systems, virtual applications, or shared services in the workload console, you might notice that the values are different than the ones that are reported on the virtual machine monitoring area in the Virtual Application Console.  

The scaling policy that is applied to a workload might seem to not be triggered, even though the scaling threshold has been exceeded.  Also, the Health Status for a virtual machine or deployment might be in Warning or Critical state, even though the displayed utilization metrics are below the defined thresholds.
ResolutionThis different in values is expected. These pages all use data that is returned from the hypervisor, which is only returning the active memory usage at any given point in time and not the allocated or in-use memory usage across a reasonable time frame.  

The Scaling Policies and Health Status for workloads use utilization values that are provided by tooling inside the guest operating system of the virtual machine, which has a more accurate view of in-use memory.  Complete the following steps to view memory values that are used for these functions:

1. Go to the Virtual Application page and click Manage to open the Virtual Application Console. 
2. In the Monitoring tab, select the virtual machine to view the utilization graphs. 

This utilization data is provided directly by the monitoring agents that are running on the operating system of the virtual machines and have access to both the guest operating system and hypervisor view of in-use memory.

Names of archive files uploaded to script packages and emergency fixes must be comprised only of ASCII characters
ProblemErrors can occur when applying emergency fixes that use uploaded archive files with names containing non-ASCII characters. Similarly, errors can occur when script packages are deployed by using archive files with non-ASCII characters. 
ResolutionBefore uploading archive files to script packages or emergency fixes, ensure that the archive file name contains only ASCII characters.

'Name or service not known' error after adding Default add NFS add-on to virtual system
ProblemWhen deploying a virtual system after adding a Default add NFS (Network File System) add-on, a 'Name or service not known' error might occur.
ResolutionBefore deploying the virtual system, specify the IP address of the NFS add-on instead of the host name in the HOST_NAME field in the Pattern Editor.

If you have already deployed the virtual system and see the 'Name or service not known' error, perform the following steps:

1. Open the workload console.
2. Click Patterns > Virtual Systems.
3. Select the pattern in the left navigation pane.
4. Click Edit to open the Pattern Editor.
5. On the canvas, locate the Default add NFS add-on, and click the edit option to open the Parameters for Default add NFS window.
6. In the HOST_NAME field, replace the host name with the IP address of the network file system.
7. Redeploy the virtual system.

Localized messages in log files for PureApplication System W1700 display unknown language characters
ProblemLocalized messages in the trace.log file contain unknown language characters. This trace.log file is associated with the installation of WebSphere Application Server after deploying the sample Web Application pattern on PureApplication System W1700.
ResolutionIgnore the unknown language characters. This is a known issue.

Certain virtual applications with 'Terminated' status cannot be deleted from the user interface
ProblemIf you click Delete to remove virtual applications in the Instances > All Instances window of the workload console, certain virtual applications that have a 'Terminated' status are not deleted, and the status continues to display 'Terminated'.
ResolutionUse the command line interface (CLI) tool to delete the virtual applications. The CLI tool is available for download on the PureApplication System Welcome page.

Note: If you plan to call IBM Support, do not delete the virtual applications first. Deleting the virtual applications also deletes all diagnostic information that is used for troubleshooting the problem.

Perform the following steps:

1. Issue the following commands to locate the virtual application that you want to delete:

pure -h <host name> -u <user> -p <password>
vas=deployer.virtualapplications["<name>"] where <name> is the name of the virtual application.
va=vas[0]
va

2. After successfully locating the correct virtual application, issue the following command to remove it.

va.delete() 

Changing the root password after deploying a virtual system instance causes problems with script packages
ProblemIf you change the root password on a virtual machine in a virtual system pattern instance, script packages that are subsequently run either manually (in when I initiate it mode) or when a virtual system is deleted do not run successfully. 
ResolutionChange the password back to the previous one, and try running the script packages again.

Cannot add users to the 'Access granted to' list for virtual applications and virtual systems
ProblemWhen working with instances of virtual applications and virtual systems in the workload console, you cannot add new users to the Access grant to field when 500 users are already in the system. 500 is the maximum amount of users that you can add through the user interface.
ResolutionUse the command line interface tool to add users beyond the 500-user limit. See the Share a virtual application with another user section in the Virtual application instance command-line interface reference topic for more information.

WebSphere Application Server virtual machines that are approaching memory limits cause errors
ProblemWhen running both the System Monitoring shared service and the System Monitoring service for WebSphere Application Server, certain WebSphere Application Server virtual machines that are approaching their memory limit in a virtual system pattern might not display in the System Monitoring Portal, or those virtual machines might display with 'Error' or 'Critical' status.

Also, the virtual system pattern deployment might display a health status of 'Running but with error (Workload agent failure)'. If you view the details for the virtual system pattern deployment, you might see a First failure data capture (FFDC) 'error=12, Not enough space' error in the agent logs in the /opt/IBM/maestro/agent/usr/servers/${servername}/logs/FFDC.log file.
ResolutionEdit the virtual system pattern to increase the memory for each WebSphere Application Server (including CustomeNode, Dmgr, and others). For PureApplication System W1500, increase the memory to 4 GB. For PureApplication System W1700, increase the memory to 6 GB. Also, check the virtual machine to confirm that it is not using too much memory.

Network issues occur in DB2 HADR environments using IPv6 
ProblemIn DB2 Enterprise Server Edition 10.5, 10.1, and 9.7 high availability disaster recovery (HADR) environments, the primary system is unable to contact the standby system using Internet Protocol version 6 (IPv6).
ResolutionThis is a known limitation. You must use IPv4 when deploying DB2 databases with HADR.

Cannot log in to the VNC web console with the virtuser account
ProblemThe password for the virtuser account must contain at least six characters. A password with fewer than six characters results in a log-in failure.
ResolutionChange the virtuser password so that it contains six or more characters.

Database deployments and cloning fail when up to 2 TB of storage space is specified 
ProblemThe initial storage size limit for AIX DB2 database deployment is 1 TB, and the limit for a Linux DB2 database deployment is 1.8 TB. If this limit is exceeded, the deployment will fail but an error message is not displayed. Cloning will also fail if the storage limits are exceeded.
ResolutionIf you want to deploy a database with a storage size of up to 2 TB, you must first create either an AIX database with a storage size no greater than 1 TB, or a Linux database with a storage size no greater than 1.8 TB. After the database is created, use the workload console to increase the storage size. If you want to clone a database deployment, the storage size for a database on AIX cannot exceed 1 TB, and the storage size for a database on Linux cannot exceed 1.8 TB.

Limitations for increasing memory for an AIX database
ProblemYou cannot scale up the memory for an AIX database beyond 4 times the size of the initial deployment.
ResolutionThis is a known limitation. If you want to deploy a database using maximum memory size of 256 GB, you must initially deploy it with 64 GB of memory.

Stopping or deleting a web application pattern results in an error without any message 
ProblemIf a user with read-only access tries to stop or delete a web applications with a database in the pattern, and error icon is displayed in the Virtual Application Patterns pane, but an error message is not displayed.
ResolutionAssign read and write access to users that you want to use for database administration tasks.

IBM Image Construction and Composition Tool



Problems using WebSphere Application Server 8.5.x virtual images in Image Construction and Composition Tool
ProblemWhen using PureApplication System W1500 and WebSphere Application Server 8.5.x virtual images in Image Construction and Composition Tool, note that the VMware ESX cloud provider does not work correctly for these images.

The following images are affected:
  • WebSphere Application Server 8.5.0.2 64-bit RHEL 6 x86-64 (VMWare)
  • WebSphere Application Server 8.5.0.2 32-bit RHEL 6 x86-64 (VMWare)
  • WebSphere Application Server 8.5.5.0 64-bit RHEL 6 x86-64 (VMWare)
  • WebSphere Application Server 8.5.5.0 32-bit RHEL 6 x86-64 (VMWare)
ResolutionFor this type of images, use the PureApplication System cloud provider instead.

Database performance monitoring



Deployed databases are not enabled for monitoring
ProblemDatabases that are deployed through a virtual system pattern before the Database Performance Monitoring shared service is started are not enabled for database monitoring.
ResolutionTo enable database monitoring for a virtual system pattern, start the Database Performance Monitoring shared service before you deploy the database.


After an HADR failover occurs, the Database Performance Monitor does not recognize the new primary database
ProblemWhen a high availability disaster recovery (HADR) primary database fails, the secondary database becomes the new primary. However, the Database Performance Monitor will not recognize the old secondary database as the new primary database.
ResolutionThis is a known limitation. After the initial primary database is restored, database monitoring will resume.

Cannot access Database Performance Monitor after an upgrade
ProblemAfter the database performance monitor is upgraded, a user opens the Database tab and sees the message "Your session has been logged out". This issue affects databases that are deployed prior to the PureApplication System 1.1.0.1 upgrade.
Resolution This is a known limitation in this release.


Page continues to load and then logs out of the session
ProblemWhen checking the status of the Database Performance Monitoring service in the workload console on the Shared Service Instances page, if you click Endpoint in the Middleware perspective section on the main details page, first a Loading message displays followed by a Your session has been logged out message.
ResolutionRestart the Database Performance Monitoring service in the workload console:
  1. Click Instances > Shared Services.
  2. Select the Database Performance Monitoring service in the list of services.
  3. Click Manage to open the Virtual Application Console.
  4. Click Operations > Database-Performance-Monitor.OPM > Restart > Submit.

'Can't find server' error when accessing advanced monitoring
ProblemAfter selecting a database on the Database Instances page in the workload console, and then clicking Manage > Monitoring > Middleware > Advanced monitoring, the browser displays a Can't find the server error message.
ResolutionRestart the Database Performance Monitoring service in the workload console:
  1. Click Instances > Shared Services.
  2. Select the Database Performance Monitoring service in the list of services.
  3. Click Manage to open the Virtual Application Console.
  4. Click Operations > Database-Performance-Monitor.OPM > Restart > Submit.


'No component exists' error even though Database Performance Monitoring service is deployed
ProblemAfter selecting a database on the Database Instances page in the workload console, and then clicking Manage > Monitoring > Middleware, a No component exists error message displays.
ResolutionRestart the Database Performance Monitoring service in the workload console:
  1. Click Instances > Shared Services.
  2. Select the Database Performance Monitoring service in the list of services.
  3. Click Manage to open the Virtual Application Console.
  4. Click Operations > Database-Performance-Monitor.OPM > Restart > Submit.

Deleted databases remain listed on the Database Performance Monitor page
ProblemWhen deleting deployed databases that are listed on the Instances > Shared Services > Database Performance Monitoring - [cloud group] > Endpoint page, note the following behaviors:

If you stop databases and then delete them, those databases remain in the user interface. This is because the databases do not get removed from the shared service registry. 

However, if you delete databases that are running, those databases are removed from the user interface because the shared service registry gets updated.
ResolutionDelete databases when they are running so that they get removed from the shared service registry. You can also manually remove databases instances that were not removed from the registry in the workload console by doing the following steps:

1. Open the workload console.
2. Click Instances > Shared Services > Database Performance Monitoring - [cloud group] > Endpoint.
3. On the Database Performance Monitor page, click the Database tab.
4. In the table, click the database connection, and click the trash can icon above the table.

Advanced Middleware Configuration



Password for automation engine client is not documented
ProblemThe root password for the Advanced Middleware Configuration automation engine is not given anywhere in the information center.
ResolutionThe password is root and should be changed immediately from within the automation engine client.

Advanced Middleware Configuration script packages are supported only on Linux and AIX
ProblemAdvanced Middleware Configuration script packages are not supported when attached to Windows based images.
ResolutionThis is a known limitation in this release.

Configuration properties files can be lost during upgrades
ProblemIf you upgrade from a version of Advanced Middleware Configuration earlier than 1.0.0.1, configuration properties files (env-config.properties) are not copied from the old location (the RAFW_HOME/work directory) to the new location (the RAFW_HOME/user/preserve/envgen directory).
ResolutionThe preferred workaround is to back up the RAFW_HOME/work directory before you upgrade, and then copy the contents to the RAFW_HOME/user/preserve/envgen directory after the upgrade. If you have already upgraded and lost your properties files, use the existing configuration option of the configuration editor to recreate the missing configuration definitions.

Custom templates fully supported for WebSphere Application Server and WebSphere Portal only
ProblemFor middleware products other than WebSphere Application Server and WebSphere Portal, you might see errors due to mismatched token names between the user-created templates and the product templates. The automated creation of custom templates is not available for the WebSphere Application Server Liberty profile.
ResolutionFor products other than WebSphere Application Server and WebSphere Portal, use product templates and customize the configuration definitions manually.

Clusters cannot recognize custom node types
ProblemNodes define what types of servers you can create on them. Clusters define what types of nodes their servers go onto. Servers are located on nodes, but are managed by clusters. Therefore, existing cluster types do not recognize custom nodes and will not allow you to create servers on them. The custom nodes are not even available in the configuration editor to add to a cluster.
ResolutionFor this release, do not use custom node types if you want servers on those nodes to be clustered. Instead, add custom questions for the standard node types.

RAFW_EnvironmentGenerationWizard automation plan no longer created
ProblemRAFW_EnvironmentGenerationWizard automation plan is missing.
ResolutionDue to extensive changes in the generation process, this plan is no longer needed. Existing copies of the plan are not deleted when you upgrade from a version earlier than 1.1, but the plan is no longer created if you install version 1.1.

User change logging suppressed in Tomcat logs
ProblemThe user change function in the automation engine is being invoked frequently enough that logging the activity was causing very large log files.
ResolutionIf you want to restore those messages, comment out the following line in the RAFW_HOME/server/tomcat/conf/logging.properties file:

com.buildforge.services.server.api.AuthContext.level=WARNING

Use a number sign (#) as the comment character.

-genCloudPattern option is no longer supported in rafwEnvBuild
ProblemThe -genCloudPattern option was removed from rafwEnvBuild.
ResolutionWith the current release, you use the Eclipse client to generate virtual system patterns (VSPs).

Default OS_RAFW_HOME changed
ProblemIn previous releases, the default value for OS_RAFW_HOME on UNIX and Linux target systems (such as AIX_RAFW_HOME) was "/opt/RAFW". With the new release, the value is now "/tmp/RAFW". The value for Windows target systems did not change.
ResolutionYou can change this default value by editing the configure.properties file on the framework server in the RAFW_HOME path.

Property name change for WebSphere Application Server Liberty profile
ProblemIf you are using a WebSphere Application Server Liberty profile configuration properties file that was created in a previous version of Advanced Middleware Configuration, you cannot generate the configuration definition from the existing file.
ResolutionYou must make a manual change to the properties file before you regenerate the configuration definition file.

Open the configuration properties file (RAFW_HOME//user/preserve/envgen/env-config.properties) in an editor. Change the name of the LIBERTY_ENVIRONMENT property to ENVIRONMENT. Change the name of the LIBERTY_PROFILE_NAME property to CONFIGURATION.

Save the properties file and regenerate the configuration definition.

Generation process for new configurations creates four automation plans
ProblemGeneration creates unexpected automation plans.
ResolutionWith this release, when you create a configuration definition from scratch, the generator not only creates the RAFW_env_config plan to create the live configuration on the target system, but also creates the same three automation plans that are generated for an existing configuration: RAFW_env_config_compare, RAFW_env_config_execute, and RAFW_env_config_import.

Problems validating the Rational Asset Manager repository URI
ProblemThe Rational Asset Manager repository URI does not require the context root for the Rational Asset Manager GUI. This is different from the Rational Team Concert repository URI which does require the context root. For example, if your Rational Asset Manager server is at ramsvt2.example.com:9080/ram/ and your Rational Asset Manager web services are ramsvt2.example.com:9080/ram.ws/RAMSecure/rootservices, then specify ramsvt2.example.com:9080/ram.ws as the value for the repository URI.
ResolutionMake sure the URIs are formatted correctly for the product you use.

Configuration not checked into source control because of "ambiguous workspace"
ProblemDue to a problem with Rational Team Concert, if you accept the default value when creating a workspace, duplicate workspaces might be created. If two or more workspaces exist with the same name, regardless of who created them, the configuration is not checked in and a message that reads "Ambiguous workspace/stream" is written to the log.
ResolutionDo not use the default value when creating a Rational Team Concert workspace.

Rational Asset Manager and Rational Team Concert integrations not available on Windows
ProblemYou can use Rational Asset Manager or Rational Team Concert to store configuration definitions for a Advanced Middleware Configuration server on UNIX or Linux systems only. The integrations with Rational Asset Manager and Rational Team Concert are not available for an Advanced Middleware Configuration server running on Windows.
ResolutionThis is a known limitation in this release.

Use "-product" option when creating Liberty profile through the command line wizard
ProblemWhen creating a configuration definition for WebSphere Liberty profile through the command line wizard, do not use the Liberty profile option in the product list. For new configurations, the definition is created in the wrong directory. For existing configurations, the wizard fails.
ResolutionStart the rafwEnvBuild command with the "-product liberty" option.

Cannot add WebSphere Application Server nodes to WebSphere Service Registry and Repository clusters
ProblemWhen you create a configuration definition for a WebSphere Service Registry and Repository cell, you can add only WebSphere Service Registry and Repository nodes to a cluster. At this time, you cannot add WebSphere Application Server or other types of nodes to the clusters.
ResolutionThis is a known limitation in this release.

Warnings in log after running integrateToBF
ProblemWhen you run integrateToBF createAll on a new installation of Advanced Middleware Configuration , you might see "Missing required parameter 'env.uuid'" warnings in the catalina log file.
ResolutionYou can ignore these warning messages.

Concurrent imports of the same cell cause deadlock
ProblemIf you have two configuration definitions based on the same WebSphere Application Server cell and you run the import automation plans from both configurations at the same time, neither plan completes.
ResolutionDo not run the import automation plans from both configurations at the same time.

Configure actions excluded for standalone servers
ProblemThe following two configure actions are not included in the was_common_configure_all action for WebSphere Application Server cells with a standalone server:

was_common_configure_health_policies
was_common_configure_service_policies
ResolutionThe problem is related to the base installation code for WebSphere Application Server. If you have a cell with a standalone server that you created using WebSphere Application Server Network Deployment installation code, you can run these two actions separately after the was_common_configure_all action completes.

False positive result of configuration storage automation plans
ProblemWhen you run one of the "get" or "put" automation plans for either Rational Team Concert or Rational Asset Manager, the plan may fail while indicating success. These plans determine their success or failure using text filters to find specific strings within the command output. If none of these strings match, then the command is deemed successful. It is possible that the command was unsuccessful, and either returned an unexpected string or possibly no string at all.
ResolutionWhen running a configuration storage automation plan, scan the output and verify that no errors were reported.

Administrative agent questions displayed for Network Deployment cells
ProblemWhen creating a WebSphere Application Server version 7.0, 8.0, or 8.5 configuration definition in the Advanced Middleware Configuration console, it is possible to specify a “Standalone” cell type, answer “Yes” for administrative agent, and then change the cell type to “Network Deployment. “ In this case, the administrative agent setting remains "Yes".
ResolutionDo not change from “Standalone” to “Network Deployment” if you have already set the administrative agent to “Yes”. If you have already made this change, delete the configuration definition and reenter the information.

Date does not change after VSP update
ProblemIf you update a VSP from a template, but do not remove the VSP from IBM Workload Deployer and reload it, your changes show up in IBM Workload Deployer, but the displayed date of the VSP does not change.
ResolutionNo workaround is required, because the VSP is updated correctly.

Generating with rafEnvBuild from older configuration properties files results in missing or incorrect values
ProblemAdvanced Middleware Configuration version 1.1 added a scope data file to the generation process. If you generate an older configuration properties file using the command line, that file is not available and errors can result.
ResolutionOpen the file in the configuration editor and generate it there.

Framework server fails to start due to missing password
ProblemIf you upgraded from Advanced Middleware Configuration version 1.0.0.3 or earlier, you might not be able to start the framework server.
ResolutionEdit RAFW_HOME/buildforge.propeties. Add a BF_ROOT_PASSWORD property and set it to the Rational Build Forge root user password. If you don't know the root password, and BF_USERNAME is set to root, you can copy the value from BF_PASSWORD to BF_ROOT_PASSWORD. Save the buildforge.properties file and restart the framework server.

was_common_configue_sibus_all does not properly update Messaging bus entries
ProblemWhen you run was_common_configue_sibus_all in execute mode, the attributes are imported, but any changes to the sib.xml file are not propagated to WebSphere Application Server when executed.
ResolutionDelete the messaging engine and recreate it with the desired datasource attributes.

Possible confusion over name of was_common_configure_security_ssl action
ProblemThe was_common_configure_security_ssl action corresponds only to Security > SSL certificate and key management > SSL configurations in the WebSphere Application Server console.
ResolutionIf you want the action that corresponds to Security > SSL certificate and key management > Dynamic outbound endpoint SSL configurations, which you use to import, execute, or work with changes, use the was_common_configure_security_global action.

WebSphere Application Server administrative console permits invalid characters in passwords
ProblemYou can include the "$" and "&" characters in a password that you create through the WebSphere Application Server administrative console. However, if you then use that password in the Advanced Middleware Configuration configuration editor, or any other process that uses a SOAP connection, the password will not be recognized.
ResolutionThe best workaround is to not create a password that uses invalid characters. You can still use a password with some invalid characters (such as "$" and "&") by prefacing them with a backslash, or by enclosing the entire password in single quotes.

The was_common_configure_start_cluster action times out
ProblemThe was_common_configure_start_cluster action times out if the cluster fails to start within 30 minutes. This can be caused by a cluster that is simply very slow to start.
ResolutionUse the was_common_configure_ripple_start_cluster action instead. This action cycles through all the servers in the cluster, stopping and restarting each one in turn so that at least one server is available at all times to handle processing.

Eclipse client uses system browser to create configuration definition
ProblemUnexpected browser used. In previous releases, the Eclipse client used an internal browser to display the Environment Generation wizard. With the current release, configuration definitions are created in the new console interface. When you click "Create Configuration Environment" in the Eclipse client, the console is displayed in the system browser rather than the internal Eclipse browser.
ResolutionNo action is required.

Creating virtual system pattern from template results in pattern for unintended part
ProblemThe VSP generator chooses parts based on the way those parts identify themselves. If, for example, a new image is an extension of a WebSphere Application Server image, but still identifies itself as the WebSphere Application Server part, Advanced Middleware Configuration  might use the new image rather than the intended WebSphere Application Server part.
ResolutionInstead of generating a VSP, use the same interface to generate a custom script package based on the template. You can then manually create a VSP in the Workload Console and add the custom script package as well as the Cleanup, Import, and Integration script packages.  Alternatively, you can edit the generated VSP, remove the incorrect image, and replace it with the appropriate one.

Virtual system pattern generation fails due to missing script package
ProblemWhen generating a virtual system pattern from a template, the following error message is displayed:

CRWSE0307E No matches for script package with name AMC Xxxxxx Script Package
ResolutionYou must have the following script packes in the IBM PureApplication System catalog:
  • AMC Cleanup Script Package
  • AMC Import Script Package
  • AMC Integration Script Package

If any packages are missing, upload them from the following location on the framework server:

/opt/IBM/RAFServer/rafw/framework/wca/

The following files are available:
  • CleanupScriptPackage.zip
  • ImportScriptPackage.zip
  • rafwScriptPackage.zip (contains AMC Integration Script Package)

After all three script packages are displayed in the catalog, generate the VSP again.


The Advanced Middleware Configuration integration with Rational Asset Manager is not available on AIX
ProblemVersion 1.1 of Advanced Middleware Configuration does not support Rational Asset Manager integration on the AIX platform.
ResolutionThis is a known limitation in this release.

Deploying an instance of Advanced Middleware Configuration from a hypervisor image fails with messages about integrateToBF
ProblemDeployment fails with the following messages.

Unable to retrieve the UUID for the entry BF_RAFW_GLOBAL_ENV_UUID in buildforge.properties. Ensure the integrateToBF script has been run. The most common symptom of this problem is later project failing to expand the MODE variable
com.ibm.rational.rafw.wsadmin.RafwException: Unable to retrieve the UUID for the entry BF_RAFW_GLOBAL_ENV_UUID in buildforge.properties. Ensure the integrateToBF script has been run. The most common symptom of this problem is later project failing to expand the MODE variable.

This is a rare condition. It is likely that Tomcat does not start and run the services for Advanced
Middleware Configuration in a timely way. This condition prevents the integrateToBF script from running, which is
required for deployment.
ResolutionDelete the instance and deploy it again. 

File name differences in /tmp directory are deliberate and should not be changed
ProblemThe default working directory of integration script package is /tmp/rafw (lower case). The LINUX_RAFW_HOME 
defined in the configure.properties file for the server is /tmp/RAFW (upper case).
ResolutionThis naming is deliberate and should not be changed.

Common messages produced when you restore by amcBackupAndRestore.sh
ProblemWhen you are using the amcBackupAndRestore.sh script to restore
Advanced Middleware Configuration, you may see the various messages.
Resolution1) These messages are produced when the database is being restored on the
new image of AMC. You can ignore them.

Restoring database 'RAFDB'...
Executing:  su - db2inst1 -c "db2 restore database RAFDB taken at xxxxxx without prompting"
SQL2540W  Restore is successful, however a warning "2539" was encountered
during Database Restore while processing in No Interrupt mode.

DB2 warning message 2539 means that a database with the given name (in this case RAFDB) already

exists and that the data in that database will be
overwritten. That overwrite is what is intended during a restore.
You can ignore this message.

2) The following messages are produced when the script is attempting to create the Advanced Middleware Configuration web UI tab. The tab is no longer used in AMC v1.1 and later releases.

Executing: /opt/IBM/RAFServer/rafw/bin/integrateToBF.sh removeUITab
Advanced Middleware Configuration for PureScale Application System version 1.1.0.1 build 20130917-1729
CRWFW0471I Invalid argument supplied:
ERROR: A valid argument was not found in the argument list.

The usage statement for integrateToBF.sh is then shown.

You can ignore this message.

Upgrading from version 1.0.x requires manual reinstallation of agents
ProblemUpgrading from version 1.0.x requires manual reinstallation of agents.
ResolutionIf you are upgrading from Advanced Middleware Configuration 1.0.x, you need to reinstall Build Forge agents manually.

1. Inside the downloaded fix pack for version 1.1.0.1, find 1.1.0.1-Rational-BFAgents-All-FP0001.zip.
2. Copy the file to a temporary directory on the deployed instance you are upgrading.
3. Unzip the file.
4. Run the agent installer for your operating system. See the information center for Build Forge version 7.1.3.5. Instructions for each operating system are in Installing agents.

If you are upgrading from version 1.1.0.0 or later, you do not need to reinstall agents manually.

Related:
List of fixes in IBM PureApplication System W1500 and W1700 Version 1.1.0.1
IBM PureApplication System release notes by version
Known issues for virtual application database patterns
Installing database pattern and database performance monitoring updates

[{"Product":{"code":"SSM8NY","label":"PureApplication System"},"Business Unit":{"code":"BU053","label":"Cloud & Data Platform"},"Component":"--","Platform":[{"code":"PF016","label":"Linux"},{"code":"PF002","label":"AIX"}],"Version":"1.1.0.1","Edition":"","Line of Business":{"code":"","label":""}}]

Document Information

Modified date:
17 June 2018

UID

swg27039551

Page Feedback