[an error occurred while processing this directive][an error occurred while processing this directive]

Preventing Excessive Media Inactivity Notifications

You can prevent a BGF from sending an excessive number, or avalanche, of media inactivity notifications to the gateway controller in a short period of time. Such an avalanche can severely degrade the performance of the gateway controller. By default, all media inactivity notifications are sent immediately from the BGF to the gateway controller. When an upstream device in the network fails, all existing terminations in the BGF report media inactivity at about the same time. You can configure H.248 notification behavior to regulate the flow of notifications sent to the gateway controller.

Regulated notification is activated either by a request from the gateway controller or by CLI commands. When you choose regulated notification you can explicitly configure the frequency for notifications of media inactivity events as follows:

  • Send only one media inactivity notification. This option is not available as part of the H.248 notification behavior package.
  • Send a configurable percentage of media inactivity notifications (0 through 100).

The Notification Behavior package only limits the sending of notifications to the gateway controller, not the occurrences of the event. We recommend you enable the recording of all media inactivity notifications for future retrieval and, optionally, request that the inactivity notifications time stamps are also enabled for future retrieval.

This topic includes the following tasks:

Configuring H.248 Notification Behavior to Prevent Excessive Media Inactivity Notifications

By properly setting H.248 notification behavior properties, you can prevent a BGF from sending an avalanche of media inactivity notifications that can flood the gateway controller and adversely affect the processing of H.248 transactions. You can also enable recording of all media inactivity events for future retrieval.

The H.248 Notification Behavior package is defined in annex E.15 of the Gateway control protocol v.3, ITU-T Recommendation H.248.1, September, 2005.

To configure default values for H.248 Notification Behavior properties:

  1. Access the configuration of H.248 application data inactivity detection.
    [edit services pgcp gateway bgf-1]user@host# edit h248-properties application-data-inactivity-detection
  2. Turn notification on.
    [edit services pgcp gateway bgf-1 h248-properties application-data-inactivity-detection]user@host# set ip-flow-stop-detection regulated-notify
  3. Access the configuration of H.248 notification behavior.
    [edit services pgcp gateway bgf-1]user@host# edit h248-properties notification-behavior
  4. Configure the default frequency of notification messages for the media inactivity event. The gateway controller can override this default by requesting a different frequency. If you specify once, only one notification is sent and the gateway controller cannot retrieve the notification by use of an audit for the root termination.
    [edit services pgcp gateway bgf-1 h248-properties notification-behavior]user@host# set notification-regulation default 10
  5. Enable retrieval of event notifications by the gateway controller at the [edit services pgcp gateway gateway-name h248-options] level.
    [edit services pgcp gateway bgf-1 h248-properties notification-behavior]user@host# up 2[edit services pgcp gateway bgf-1]user@host# edit h248-options[edit services pgcp gateway bgf-1 h248-options]user@host# set audit-observed-events-returns
  6. Enable retrieval of time stamps of recorded notifications by the gateway controller.
    [edit services pgcp gateway bgf-1 h248-options]user@host# edit services pgcp gateway bgf-1 h248-properties[edit services pgcp gateway bgf-1 h248-properties]user@host# set event-timestamp-notification request-timestamp requested

Managing the Rate for All Notifications Sent by a PIC or DPC

You can limit the rate at which the PIC or DPC sends all notifications. Using this functionality, you can prevent an avalanche of media inactivity notifications and throttle all notifications to a configured rate. Doing this enables the system to maintain a stable state when the PIC or DPC is generating a large volume of messages.

You can configure the aggregate rate of notifications coming from a PIC or DPC. Using this approach, you limit the rate for all messages from the PIC or DPC, not just notifications.

  1. Access the configuration of PGCP parameters.
    [edit services]user@host# edit pgcp
  2. Configure the rate (per second) for PICs to send messages to the gateway controller.
    [edit services pgcp ]user@host# set notification-rate-limit25

Published: 2010-04-13

[an error occurred while processing this directive]