Lucent Technologies 101-1620-005 Manuale Utente

Pagina di 120
Octel Unified Messenger Concepts & Planning Guide
2-4
101-1620-005
Sizing ports
Accurate port sizing is critical to the successful implementation of any 
voice processing application. A careful analysis of system traffic patterns 
must be made. Specifically, the volume of calls generated by internal 
subscribers and external callers to the system during the busiest hour of 
the day must be determined. 
Busy hour traffic is used in the calculation to determine the number of 
ports needed to provide an acceptable level of service to the users. Most 
newer PBXs provide traffic statistic reports that can provide an accurate 
picture of traffic patterns on the switch. When preparing the 
implementation of a messaging system, the planner should study a 
minimum of one week’s traffic data to determine daily and hourly call 
volumes. 
How to determine 
the busy hour
Most PBXs can generate traffic reports that provide statistics on a weekly, 
daily, or hourly basis. Usually these reports break out the traffic statistics 
by type of call, for example, incoming calls, outgoing calls, and calls to 
specific hunt groups. These reports can be used to determine specific 
traffic patterns and when the busy hour occurs. 
Suppose that during the course of a typical business week the traffic 
report indicates that Monday is the busiest day for incoming calls. On 
Monday a total of 3,500 calls are received, and the hourly statistics show 
that the busiest hour of the day occurs between 9 a.m. and 10 a.m., with 
350 incoming calls. Therefore 9 a.m. to 10 a.m. is the customer’s busy 
hour.
Another way of expressing busy hour traffic is as a percentage of total 
daily traffic. In the example above, 350 busy hour calls is 10% of the 
3,500 total calls received for the day. If traffic statistics are not available, 
then an educated guess at busy hour traffic is required. The tables in 
Appendix A, “Sizing for Ports” can be used as guidelines for estimating 
how many users can be supported with a given number of ports based on 
some assumptions about average daily port usage per subscriber. These 
tables also provide for traffic patterns of 10%, 14%, or 18% busy hour 
traffic.