Cisco Cisco Prime Unified Operations Manager 8.7 White Paper
Operations Manager Deployment Best Practices
All contents are Copyright © 1992–2006 Cisco Systems, Inc. All rights reserved.
Page 84 of 112
example. You do not need to generate a trap and be alerted for every call, only for the
calls that are experiencing poor quality.
Instead of using the archived files, you can use the Operations Manager GUI (when
Operations Manager and Service Monitor are co-resident), to assist in determining your
QoV baseline. In Service Monitor, configure the Service Monitor MOS threshold to 4.5
(for a short period of time) to determine a QoV baseline for your network. This will
generate a QoV trap for every call. Monitor the MOS and determine the typical QoV
MOS for your network. Then change the Service Monitor MOS to a lower permanent
value based upon your determined baseline.
calls that are experiencing poor quality.
Instead of using the archived files, you can use the Operations Manager GUI (when
Operations Manager and Service Monitor are co-resident), to assist in determining your
QoV baseline. In Service Monitor, configure the Service Monitor MOS threshold to 4.5
(for a short period of time) to determine a QoV baseline for your network. This will
generate a QoV trap for every call. Monitor the MOS and determine the typical QoV
MOS for your network. Then change the Service Monitor MOS to a lower permanent
value based upon your determined baseline.
Instead of using the archived files, you can use the Operations Manager GUI to assist in
determining your QoV baseline. You can do this provided that you have configured
Service Monitor to send traps to Operations Manager and that you have added Service
Monitor to Operations Manager. To determine the QoV baseline:
determining your QoV baseline. You can do this provided that you have configured
Service Monitor to send traps to Operations Manager and that you have added Service
Monitor to Operations Manager. To determine the QoV baseline:
1. In Service Monitor, configure the Service Monitor MOS threshold to 4.5. As a
result, Service Monitor will generate a trap for every call, and Operations
Manager will generate alerts.
Manager will generate alerts.
2. In Operations Manager, launch the Service Quality Alerts display; monitor the
MOS and determine the typical QoV baseline MOS for your network.
3. In Service Monitor, configure the MOS threshold to a lower permanent value
based upon your determined baseline.
4. In Operations Manager, verify that the MOS threshold for Critical Service Quality
Issues (in Service Quality Event Settings) is set lower than the MOS threshold in
Service Monitor.
Service Monitor.
7.6.2 Cisco 1040 Sensor in Sampling Mode
Cisco 1040 Sensors are capable of monitoring 100 RTP streams. If a Cisco 1040 sensor is
deployed on a switch that has more than 100 RTP streams, the sensor will perform
sampling, in which case, some of the RTP streams will not be considered for MOS value
generation. This situation must be avoided at all times.
In sampling mode, the reported MOS value is diluted, because some of the RTP streams
are not considered. The sensor monitors RTP streams and collects the information
necessary to compute the MOS value. This information is stored in a buffer from which
the computation process obtains the data to compute the MOS value. If packets arrive at a
rate faster than the buffer can be emptied, some of the RTP streams will be dropped
before the sensor collects information from them.
We have to keep in mind that CPU resources will be utilized constantly. Hence, it is not
just the buffer that becomes the bottleneck when a sensor is overwhelmed with excess
RTP streams, but also the CPU falls short in serving the different processes. The MOS
value reported by the sensor gets diluted as the number of simultaneous RTP streams
deployed on a switch that has more than 100 RTP streams, the sensor will perform
sampling, in which case, some of the RTP streams will not be considered for MOS value
generation. This situation must be avoided at all times.
In sampling mode, the reported MOS value is diluted, because some of the RTP streams
are not considered. The sensor monitors RTP streams and collects the information
necessary to compute the MOS value. This information is stored in a buffer from which
the computation process obtains the data to compute the MOS value. If packets arrive at a
rate faster than the buffer can be emptied, some of the RTP streams will be dropped
before the sensor collects information from them.
We have to keep in mind that CPU resources will be utilized constantly. Hence, it is not
just the buffer that becomes the bottleneck when a sensor is overwhelmed with excess
RTP streams, but also the CPU falls short in serving the different processes. The MOS
value reported by the sensor gets diluted as the number of simultaneous RTP streams