VMware

VMware vFabricTM Application Director Release Notes

vFabric Application Director 5.2 | 30 May 2013 | Build: 1052405

Last Document Update: 30 May 2013

These release notes cover the following topics:

What's New

VMware vFabric Application Director is an application provisioning, update life cycle, and release automation solution that simplifies the creation and standardization of application deployment topologies across hybrid cloud services. This release includes the following new features.

Provision and Update Life Cycle Support on vSphere
  • Deploy Linux and Windows-based applications on vSphere by leveraging vCloud Automation Center as the Cloud Abstraction Layer
  • Scale existing deployed applications on the vSphere infrastructure
  • Update the code and configuration of existing deployed applications on the vSphere infrastructure
Windows Update Support
  • Update Windows-based applications on the vCloud Director and vCloud Automation Center infrastructures
  • Native support for BeanShell, PowerShell, and CMD scripts to enable Windows-based application deployments

Other vFabric Application Director Documentation

vFabric Application Director 5.2 includes the following product documentation.

Known Issues

Installation

  • When instantiating the vFabric Application Director vApp to your cloud from the vCloud Director catalog, changing the default full name at the Configure Virtual Machines step causes the vFabric Application Director vApp to fail to start up.
    You see an error message similar to the following,
    • Unable to start vApp "My-ApplicationDirector".
    • Invalid vApp properties: Argument 'none' not a valid network name.
    Workaround: If you must replace the full name of the vFabric Application Director virtual machine, edit the properties of the vFabric Application Director virtual machine after the instantiation process completes.

Cloud Abstraction

  • When creating a custom virtual machine template for vCloud Automation Center, make sure that all of the network configuration artifacts are removed from the network configuration files.
  • When creating a cloud provider for vCloud Automation Center, if you type an invalid IP address or host name you receive the following error messages,
    • An error occurred in the cloud: An error occurred in the cloud: no protocol: 192.0.2./
    • An error occurred in the cloud: An error occurred in the cloud: no protocol: test-vcac-dev.com/
    Workaround: Type a valid and secure vCloud Automation Center IP address or host name URL, such as https://192.0.2.115 or https://test-vcac-dev.com.
  • You cannot update the host IP address of a cloud provider that has the vCloud Automation Center provisioning group and administrator credentials for the provisioning group already configured.
    Workaround: Create another vCloud Automation Center cloud provider and add the new host IP address.
  • Sometimes, when creating a deployment environment for Amazon EC2, us-west-1b might be designated as the Availability Zone. Ignore the option us-west-1b and type the applicable Availability Zone for your Amazon VPC.

vFabric Application Director Catalog

  • In some cases, installation of the vFabric tc Server service might fail when you try to download binaries from the default mirror site because of network connectivity or other intermittent problems.
    Workaround: Repeat the installation process. You can also configure the service INSTALL script to use another mirror site.
  • In the Clustered DotShoppingCart sample application, changing the default Apache load balancer http_proxy_port property value in the blueprint and deploying the sample application, causes the application to become inaccessible from a Web browser.
  • For the jPetStore 1.0 sample application, if you override the default tc Server instance_name, tc_install_path, SQLfire client_port, and sqlfire_install_path property values in the deployment profile during deployment, the deployment is successful but operations in the jPetStore application fail.
  • When authoring a new service, you cannot install many applications and RPM packages with SELinux enabled.
    Workaround: Disable SELinux by default when creating new templates. When using a sample template for download, you must disable SELinux as part of the service install script. The following example shows how you can disable SELinux on a CentOS operating system.
    if [ -x /usr/sbin/selinuxenabled ] && /usr/sbin/selinuxenabled; then
     if [ -x /usr/sbin/setenforce ]; then
      /usr/sbin/setenforce Permissive
     else
      echo 'SELinux is enabled. This may cause installation to fail.'
     fi
    fi
  • Before you deploy a vFabric Gemfire 6.6.0 service in an environment with a proxy server, update the proxy arguments for the RPM install command in the service install script. The following example shows the proxy argument values for the RPM install command.
    rpm --httpproxy proxy.vmware.com --httpport 3128 -Uvh --force http://repo.vmware.com/pub/rhel${REV}/vfabric/${VERSION}/${REPO_RPM}
    Replace the HTTP proxy host and HTTP proxy port values.
  • During the installation of the sample RabbitMQ service on Ubuntu, the apt-get command might intermittently fail with incorrect checksums depending on the network infrastructure.
    Workaround: Redeploy the application deployment or contact your system administrator if the failure persists.

Application Modeling

  • If there is a cyclic dependency between nodes in an application blueprint, the execution plan is rendered as blank in the deployment profile without displaying an error message.
    Workaround: Remove the cyclic dependency between the nodes and deploy the application.
  • vFabric Application Director assigns the appropriate memory values for an Amazon EC2 instance. Manually configuring the memory settings might cause a failure in deployment.

Application Deployment and Update Process

  • If you use the SSH client to log in to an application deployed to vSphere using vCloud Automation Center, power off a virtual machine in the application from the vCloud Automation Center or vCenter Server user interface, and initiate an update process, the process remains in progress indefinitely.
    Workaround: Cancel the update process, power on the virtual machine in the application, and initiate another update process.
  • The VM Details status window on the deployment summary page might render blank without an error message.
    Workaround: Refresh your Web browser.
  • Sometimes, when the vFabric Application Director server times out, the Web interface displays the error message, Unable to communicate with server. Please contact the administrator.
    Workaround: Reopen the application or deployment.
  • If you shut down the vFabric Application Director appliance while application deployments are in progress and restart later, the interruption might not allow the deployment status details to be updated properly.
  • In some cases, when a deployment task fails, tasks that are running in other nodes display an In progress status infinitely even though the tasks have completed successfully.

User Interface

  • Sometimes, after you close an error message box in the Web interface you might not be able to reopen it.
    Workaround: Close the browser and open the vFabric Application Director appliance in a new browser.
  • If an application has more than 12 application versions, you might experience problems when navigating between the application versions.
  • When attempting to perform copy and paste operations to define an action script for a property or attempting to paste a property value, the Adobe Flash-based user interface page might become unresponsive.
    Workaround: Make sure that you are running the supported Adobe Flash Player 11.2 or later in your Web browser. Copy the script text into an application such as Notepad and paste the text to the Web interface.

Top of Page