VMware

VMware vFabric Data Director 2.5 Release Notes

VMware vFabric Data Director 2.5 | 6 December 2012 | Build 927633

VMware vFabric Postgres 9.1.6.0 | 16 October 2012 | Build 852861

Check these release notes for additions and updates.

What's in the Release Notes

These release notes apply to vFabric Data Director Release 2.5 and discuss the following topics:

What's New in vFabric Data Director 2.5

vFabric Data Director is a data virtualization platform that enables provisioning and lifecycle management of Oracle, Microsoft SQL Server, vFabric Postgres, and Hadoop on VMware vSphere.

Data Director 2.5 expands the platform to include support for Microsoft SQL Server 2008 R2, Microsoft SQL Server 2012, and major Hadoop distributions. The 2.5 release also expands the operating system support for Oracle, Redhat Enterprise Linux version 5.4 and above, and all corresponding Oracle Linux versions, and Windows Server OS.

  • Microsoft SQL Server support.
    • Support for SQL Server 2008 R2 and SQL Server 2012.
    • Integration of SQL Server DBVM with Active Directory.
  • Support for Hadoop provisioning for all major distributions.
    • Deploy a Hadoop cluster on vSphere quickly and easily.
    • Ability to fully customize the configuration the profile to meet your needs.
    • Manage and monitor Hadoop easily.
    • Increase and decrease compute capacity on demand.
    • Improves availability of Hadoop clusters.
  • Automation of SQL Server and Oracle template creation. Create a template from scratch by providing ISOs for the OS and the database.

  • Broad support for Redhat Enterprise Linux and Oracle Enterprise Linux. Support for Redhat Enterprise Linux (RHEL) 5.4 and later (6.0 is not supported); Oracle Linux 5.4 and later (5.7 and 6.0 are not supported).

  • Enhancements to Oracle database ingestion, including ingestion to a point-in-time.
    • Use in-place refresh to maintain the OS and database configuration after refreshing.
    • Ingest a one-time clone or in-place refresh of an external database at a specified point-in-time.
    • Ingest or in-place refresh to an external database.
    • Ingest or in-place refresh to a point-in-time between two backup sets.
  • Support for Static IP for DBVMs.
  • Express setup to install Data Director. Express setup enables you to quickly create a simplified Data Director deployment for development or experimentation. Express setup creates a Data Director deployment using a single VLAN and a set of predefined system resources, organizations, and database groups.

Installation Notes for This Release

Read the vFabric Data Director Installation documentation for step-by-step instructions on installing and configuring Data Director. Before you start the installation process, use the vFabric Data Director Deployment Worksheets to help you plan your Data Director configuration.

Upgrades for This Release

A vFabric Data Director upgrade might involve two tasks: upgrading the Data Director virtual appliance and upgrading the database virtual machines (DBVMs). A particular upgrade might require that you upgrade only the Data Director virtual appliance or that you upgrade the database virtual machines as well. The Data Director virtual appliance supports in-place upgrades from all versions directly to 2.5. Use VMware vSphere Upgrade Manager to upgrade the Data Director virtual appliance. See the vFabric Data Director Installation documentation for information on upgrading the database virtual machines.

For information about upgrading using vSphere Upgrade Manager, see the following documentation.

Notes for the Product Guides

The following information is not currently addressed by the product guides.

  • When repairing a database, the available database storage allocation must be equal to or greater than the storage used by the database. This is because the repair procedure clones the database before repairing it, and there must be enough storage available to create a temporary additional copy of the database.

  • You can repair a database regardless of database status. The status for each database may change as it is being repaired. The final status is: Running.

  • To delete a user you must first perform the disable operation on that user.

Known Issues

Data Director 2.5 has the following known issues. If you encounter an issue that is not in this known issues list, search the VMware Knowledge Base, or let us know by contacting VMware Technical Support.

  • In the vFabric Data Director for Hadoop user interface, Cluster create fails with no error message to indicate the log location.
    In the user interface, if the datastore, resource pool or network is not correctly configured, the cluster create operation fails and no error message is displayed.

    Workaround: Attempt to create a cluster in the Serengeti CLI and examine the error message to determine the cause of the cluster create failure. You can do this from the server or from the client.

    Create a cluster in the Serengeti CLI from the server

    1. Use Secure Shell (SSH) to connect to the server and log in as serengeti/password.
    2. Type the command: serengeti to launch the Serengeti CLI.
    3. Connect to the server by typing the command: connect --host localhost:8080 --username serengeti --password password.
    4. Type the command: cluster create --name FAILED_CLUSTER_NAME --resume
    5. The CLI output error message will be similar to the following, indicating the log location for the failure.

      FAILED 15%

      cluster default73 create failed: create failed with: Can not alloc resource for vm group master: Input available host list cannot be empty. you can get task failure details from serengeti server log at: /opt/serengeti/logs/task/1

    6. Determine the cause of the cluster creation failure and fix the problem.
    7. Type the command: cluster create --name FAILED_CLUSTER_NAME --resume

    Create a cluster in the Serengeti CLI from the client

    1. From the Download page, download the ZIP file containing the CLI console by clicking the link: Download vFabric Data Director for Hadoop CLI console.
    2. Unzip the file.
    3. Ensure that Java is installed correctly.
    4. Open a command window and make the upzipped directory your working directory.
    5. Type command: java -jar serengeti*.jar to launch the Serengeti CLI.
    6. Connect to the server by typing the command: connect --host SERENGETI_SERVER_NAME:8080 --username serengeti --password password
    7. Type the command: cluster create --name FAILED_CLUSTER_NAME --resume
    8. The CLI output error message will be similar to the following, indicating the log location for the failure.

      FAILED 15%

      cluster default73 create failed: create failed with: Can not alloc resource for vm group master: Input available host list cannot be empty. you can get task failure details from serengeti server log at: /opt/serengeti/logs/task/1

    9. Determine the cause of the cluster creation failure and fix the problem
    10. Type the command: cluster create --name FAILED_CLUSTER_NAME --resume

  • Cannot use CMS PITR in vSphere 5.1
    For vSphere 5.0, vFabric Data Director supports Central Management Server (CMS) Point-in-Time Recovery (PITR) restores using VMware Data Recovery. For vSphere 5.1, VMware Data Recovery is replaced by VMware Data Protection (VDP), which supports only virtual machine level backups. This means you cannot perform a CMS PITR restore automatically.

    Workarounds: Choose one of the following options.

    • Use VDP to back up the CMS daily to minimize data loss. VDP, based on EMC Avamar, is very good in deduplication. You can perform backups at fairly high frequency.

    • Before performing a CMS restore, copy the CMS meta database's archived log to a Network File Server. After the CMS is restored, copy the archived log back, then run a script to restore the meta database to some point in time.