Intelligent polling for multi switch networks
1. wait time 0 secs hearing ringback
2. consider skill 1 pri m adjust-by 0
3. consider skill 2 pri m adjust-by 20
4. goto step 20 if expected-wait for best = 0
5. consider location 1 adjust-by 30
6. consider location 2 adjust-by 30
7. consider location 3 adjust-by 30
8. consider location 4 adjust-by 30
9. consider location 5 adjust-by 30
10. consider location 6 adjust-by 30
11. consider location 7 adjust-by 30
12. consider location 8 adjust-by 30
13. consider location 9 adjust-by 30
14. consider location 10 adjust-by 30
15. consider location 11 adjust-by 30
16. consider location 12 adjust-by 30
17. consider location 13 adjust-by 30
18. consider location 14 adjust-by 30
19. consider location 15 adjust-by 30
20. queue-to best
21. announcement 1001
22. wait time 60 secs hearing music
23. goto step 21 if unconditionally
First, even in very large networks you can obtain nearly all of the possible benefits in agent utilization with very few polling connections. In a network of 16 switches, 99 percent of the total benefits possible with BSR can be obtained if each switch polls just 4 others.
Now our vector looks like the following. Is polling time now cut from 15 seconds to 4 seconds, proportional to the reduction in consider steps?
1. wait time 0 secs hearing ringback
2. consider skill 1 pri m adjust-by 0
3. consider skill 2 pri m adjust-by 0
4. goto step 9 if expected-wait for call = 0
5. consider location 5 adjust-by 30
6. consider location 10 adjust-by 30
7. consider location 13 adjust-by 30
8. consider location 15 adjust-by 30
9. queue-to best
10. announcement 1001
11. wait time 60 secs hearing music
12. goto step 10 if unconditionally
In fact, polling time in this vector can be around 0.4 seconds per call because of mechanisms in BSR that constantly react to network conditions and resource usage to minimize the number of status polls. The mechanisms, whose combined operation is called Intelligent Polling
, also function to make each status poll as productive as possible.