Table of Contents
Advertisement
All contents are Copyright © 1992–2007 Cisco Systems, Inc. All rights reserved. This document is Cisco Public Information.
TCP/UDP Ports Required During a Failover
Communication between Stream Manager devices requires a network that is enabled for IP
multicast. The "Matrix-based N+N Redundancy" section on Page 3-1 explains the steps that occur
when a failure is detected. This section describes the network communication and the TCP/UDP
ports that are used during a failure.
Figure 75 shows the different TCP/UDP ports used during a failover event.
Figure 75.
Example—Failover ISP
The Stream Manager Administration and Monitoring with Failover Module must be able to
communicate with all Stream Manager devices in the environment. This module polls a device
three times before logging a failure. In this example, ISP_3 backs up video streams that are
recorded by ISP_2.
1.
The Administration and Monitoring with Failover Module is configured to poll devices every 20
seconds, sending Discovery Requests and Failover Discover Requests (target UDP port 4111)
on multicast group 235.1.1.1. These requests include a new target UDP port (range 4300–
4500) in which devices must respond. Active Stream Manager devices respond with
information such as device type, name, and configuration of each camera. The Administration
and Monitoring with Failover Module reviews the responses of all active devices and
determines what ISPs are configured as failover ISP.
2.
When ISP_2 fails to respond three times, the Administration and Monitoring with Failover
Module initiates a new TCP session with ISP_3 using the failover port (TCP port 4117) and
communicates the camera ID (or peripheral ID) of the first port to back up.
3.
Via the serial cable, ISP_3 instructs the matrix switch to switch video from the failed port to the
first configured monitor output port (Camera ID# to monitor Output# ). ISP_3 begins to record
on its input port.
Design Guide
III
Page-III
Advertisement
Table of Contents
