Jump to: navigation, search

Voice Platform Resource Manager Release Notes

Release Date Release Type Restrictions AIX HP-UX PA HP-UX IPF Linux Solaris Windows
03/23/15 General X X

Helpful Links

New in This Release

There are no restrictions for this release. This section describes new features that were introduced in this release of Voice Platform Resource Manager.

  • Resource Manager can now be configured to reject a call request if the geo-location of the targeted Logical Resource Group (LRG) does not match with geo-location attribute of the call request. This extends to all calls, a behavior that previously applied only to recording solution calls. Read about this behavior and the option that controls it, reject-on-geo-location-nomatch (new in release 8.5.1), in Locating Resources Using Geo-Location.
  • A new option for the Resource Manager application enables releasing the connection when the ActiveMQ Producer sends a method that is blocked due to network disconnect or congestion issues.
    Section: ems
    Option: rc.amq_connection_send_timeout
    Valid Values: Any integer greater than or equal to 45.
    Default Value: 60.
    Takes effect: After start/restart. When this option is set to true, Resource manager waits for the time specifies in AMQ Thread Connection Send Timeout and then releases the connection.

Corrections and Modifications

This release also includes the following corrections or modifications:

Resource Manager no longer freezes when physical resources in an LRG are dynamically updated. Previously, simultaneous SNMP and dynamic updates triggered a specific race condition, and then the freeze. (GVP-22119)

Resource Manager memory no longer grows excessively when connected to SNMP Master Agent. (GVP-22078)

In a High Availability Active-Active configuration, Resource Manager (RM) now successfully records the call even when the recording request first goes to an MCP that is gracefully shutting down. It locates another MCP and notifies the other node that the new MCP must be used for the conference. Previously, the other RM node continued to use the MCP that was shutting down, which caused the call recording to fail. (GVP-21899)

Upon receiving a SUBSCRIBE request with a TO-tag, Resource Manager now returns the message 481 Call Leg/Transaction Does Not Exist if the subscription does not exist. (GVP-21881)

When Resource Manager is in active-active mode, calls no longer fail during this race condition scenario: treat-campaign-as-conference is true for OCS calls, and calls with the same campaign ID arrive on both of the nodes simultaneously. (GVP-21815)

In a High Availability Active-Active configuration, the Resource Manager with member id=2 no longer terminates at start-up if it is immediately started after the Resource Manager with member id=1 is started. (GVP-21586)

In a high availability active-active configuration, the Resource Manager no longer marks physical resources that have monitor-method set to options, as offline instead, when it comes up. Previously, this default behavior caused new calls to an existing conference to be rejected by the newly brought up RM. (GVP-21560)

A Resource Manager in a high-availability, active-active configuration will not terminate upon receiving a conference call, if both nodes in the configuration use the same hardware address. (GVP-21065)

This page was last edited on April 29, 2015, at 23:57.
Comments or questions about this documentation? Contact us for support!