vFabric Administration Server 1.1 Release Notes

vFabric Administration Server 1.1.1 | 06 DEC 2012
vFabric Administration Server 1.1.0 | 16 OCT 2012

Last Document Update: 7 JAN 2013

What's in the Release Notes

The release notes cover the following topics:

What's New in vFabric Administration Server 1.1.1

vFabric Administration Server release 1.1.1 is a bug fix release. See Known and Resolved Issues for issues resolved in the 1.1.1 release.

  • The release includes improvements and corrections in the REST API documentation. See the updated REST API documentation for the 1.1.1 release at http://www.vmware.com/support/developer/vas/rest-api-1.1.1.RELEASE.
  • The fix for VMS-1199 adds a new feature, the ability to configure the environment for the administration-broker and administration-server scripts. You can create an administration-broker.env and administration-server.env file in either the $VAS_HOME/bin or /etc directories. The .env file can contain shell commands to set the environment, for example, adding the erl command to the PATH.

What's New in vFabric Administration Server 1.1.0

vFabric Administration Server release 1.1.0 contains the following changes and new features:

  • vFabric Administration Server 1.1 administers vFabric SQLFire and vFabric Web Server on Linux nodes. REST APIs and Python and Ruby bindings have been added for SQLFire and Web Server.
  • The state of a group instance is a roll-up of the states of the individual nodes in the group. If the nodes do not all have the same state, the state is MIXED. The state of a revision deployed on a vFabric tc Server group instance is also a roll-up value, and can be reported as MIXED.
  • The 1.1 release includes new node metadata features that allow you to customize configuration files on each node in a group instance. Customized node metadata is set through the REST APIs, or by editing a properties file on each node.

    Replaceable tokens, called eyecatchers, can be embedded in configuration files. The node-specific values to replace eyecatchers are stored as metadata on each node. When a configuration file is distributed to a group instance, the vFabric Administration Server agent on each node replaces eyecatchers with values retrieved from the node's metadata.

    The vFabric GemFire and vFabric SQLFire REST APIs for locator-instances also use node metadata properties to look up the network addresses that locators bind to on each node.

    Read more about node metadata in vFabric Administration Server Concepts.
  • In addition to the ip-addresses field that contains a list of both IPv4 and IPv6 addresses, the node details data now include ipv4-addresses and ipv6-addresses fields so that IPv4 and IPv6 addresses can be accessed separately.
  • As with each new release of vFabric Suite, if you use Red Hat Enterprise Linux (RHEL), you install a new VMware repository configuration RPM. This new installation enables you to easily browse and install the vFabric component RPMs associated with vFabric Suite 5.2, such as vFabric Administration Server 1.1. In addition, the 5.2 repository RPM installation now asks you immediately to accept the End User License Agreement (EULA). In previous releases, you accepted the EULA the first time you installed a vFabric component associated with the Suite release. See RHEL: vFabric Administration Server from an RPM.
  • The VMware RPM repository vfabric-all is deprecated and will no longer be updated with new RPMs. In addition to vFabric Suite components, vfabric-all contained releases of vFabric components that were not associated with a vFabric Suite release. If you want to install a vFabric component that is not yet part of a vFabric Suite release, you must download the RPM from the VMware Download Center and install it using rpm -ivhf. Under RHEL: vFabric Administration Server from an RPM, see "Install vFabric Administration Server from a Downloaded RPM."

vFabric Administration Server REST API documentation for this release can be found at https://www.vmware.com/support/developer/vas/rest-api-1.1.0.RELEASE/index.html.

Known and Resolved Issues

The following problems were identified in vFabric Administration Server. The table indicates the version in which the problem was found and, where applicable, the version in which it was fixed. If the Fixed In column is blank, it means the problem still exists in this version of vFabric Administration Server.

Issue Number Description Found In Fixed In
VMS-1171 When a vFabric GemFire locator group instance is created, the address on which the locator listens can be configured. This address is the address of a network adapter and, because every node has a different IP address, the address needs to differ from node to node. Configuring it directly on a group instance does not work, as that would result in the same address on every node. Instead you specify the address as a property key. On each node, the value is looked up in the agent's agent.properties file—the node metadata. If the property is not specified or does not exist in the metadata, then vFabric GemFire by default binds to the default network adapter. This fix takes advantage of the new node metadata feature introduced in release 1.1.0 that enables node-specific configuration. 1.0.0 1.1.0
VMS-1175 Setup fails on Windows if the installation path contains spaces. Updating the embedded tc Server to 2.8.0 (VAS 1.1.1) and 2.7.2 (VAS 1.0.2) fixes this. 1.0.0 1.1.1
VMS-1199 "erl: command not found" when starting the Administration Broker and Server, because the erl command is not in the path. The path at execution time may differ depending on whether the broker and server are started at the command line or as services. To ensure the path is correct when the broker and server run, create files $VAS_HOME/bin/administration-broker.env and $VAS_HOME/bin/administration-server.env (or /etc/administration-broker.env and /etc/administration-server.env) and update the path in them. For example, if the erl command is in /usr/local/bin, EXPORT PATH=$PATH:/usr/local/bin. 1.0.1 1.1.1
VMS-1213 The same node may be registered twice with the server if its announcement message is redelivered. 1.0.0 1.1.1
VMS-1214 Creation of a vFabric RabbitMQ instance is not idempotent and fails if a message is redelivered. 1.0.0 1.1.1
VMS-1215 Creation of a vFabric tc Server instance on Windows is not idempotent and fails if a message is redelivered. 1.0.0 1.1.1
VMS-1216 Deletion of an instance may race with a log state update, causing a benign NodeInstanceNotFoundException to appear in the server's and the agent's logs. 1.0.0 1.1.1
VMS-1221 When a configuration is deleted or its contents are updated, the server's copy of the old contents is not cleaned up correctly. 1.0.0 1.1.1
VMS-1224 If a 32-bit JVM is running on a 64-bit OS, a node incorrectly reports its architecture as 32-bit. 1.0.0 1.1.1
VMS-1225 Documentation of /web-server/v1/nodes/{}/instances/{}/configurations/live/{}/ does not list a content link. 1.1.0 1.1.1
VMS-1229 Directory entries in a group archive zip are not created properly. 1.1.0 1.1.1
VMS-1237 Creating an installation using a tc Server compressed tar archive (tar.gz) fails. The workaround is to use the tc Server ZIP installation archive. 1.0.0