I haven't seen a max, but yes as you adjust the polling ratio to be less frequent you are freeing up resources that would let you handle more elements.
At the defaults I have seen a poller using only NPM handle 15k elements and was sitting right around 79% usage. So in theory you could poll more by turning down the polling interval, but I don't know if you would be able to eclipse 100k elements, or if there is a hard limit at some point.
I haven't seen anyone ever go this high. I believe one client I had was up to 24k on one poller, but they didn't adjust the polling ratio, and the poller was overloaded and dynamically adjusting the ratio itself, but it was still losing polling information, and not reaching the 99 - 100% completion ratio. This was an oversight by there part and we adjusted the elements by load balancing them between pollers.
Going to high in your polling interval really starts to become useless since you aren't going to be very proactive to issues. In this scenario you won't know about up/down status until your 20 minute polls go off.
Thanks for quick response!
I should have said goal is to poll switch user access ports and create reports showing distinct interfaces used during any 8 hour period to explain this unusual question.
We initially thought about using NPM default polling and UDT but at first glance it appeared we would need 1 or 2 NPMs and 8 additional pollers.
So we were wondering if we could create same reports by retarding/slowing polling frequency and tracking interface availability.