You do not have to poll every Communication Manager in larger networks. Because of the intelligent polling capabilities of BSR, you can obtain 99 percent of the possible benefits in agent utilization with very few polling connections.
The following example is a laboratory network of 16 Communication Manager used for simulations of BSR multisite applications.
As shown in the following table, approximately 99 percent of the possible benefits are obtained when any one Communication Manager polled four others.
Number of remote sites polled by each switch |
ASA across the network (seconds) |
Approximate percentage of total benefits obtained (%) |
0 |
192.8 |
0 |
1 |
26.2 |
89 |
2 |
10.6 |
95 |
3 |
7.6 |
98 |
4 |
6.5 |
99 |
15 |
4.7 |
100 |
For each Communication Manager to poll the other 11 Communication Manager in the network produces an additional percent gain in ASA and agent utilization, an improvement which is more than offset by the cost of additional messaging and trunking.
In most situations, you can get the optimal results with your multisite BSR applications if you follow the polling guidelines shown in the following table.
Number of switches in the network |
Number of switches to poll |
2-4 |
all the other switches |
5-10 |
3 other switches |
11-20 |
4 other switches |
21-40 |
5 other switches |
More than 41 |
6 other switches |