How Much Polling is Enough?  

With more and more riding on a business’s ability to have its enterprise network perform at a high level, network managers continue to push the envelope in terms of polling frequency for network assets.  Especially in larger, more complex networks, network operators are pressured to poll literally tens of thousands of devices at less than one minute intervals!  Beyond the obvious challenges in managing the sheer amount of data involved, there is a significant amount of inefficiency in collecting all that data that can be overcome with a more intelligent (i.e. “stateful”) polling architecture.

NerveCenter’s Architecture can Help.  

NerveCenter’s unique finite-state architecture can help solve the challenges with high-frequency polling, especially in very large networks.  NerveCenter allows network operators to apply custom logic to intelligently and automatically trigger polls only when it’s necessary to collect the relevant information, which by design is custom to an operator’s own unique requirements.

• NerveCenter can be set to poll at relatively large intervals until or unless some custom sequence of events occurs over time, whereby NerveCenter would automatically poll more frequently for a certain limited subset of network assets.
• Today, current customers are using NerveCenter in very large scale networks, polling literally hundreds of thousands of IP addresses with one NerveCenter server.

NerveCenter’s efficient, flexible and intelligent polling can collect only the necessary information, which significantly reduces the overhead, complexity and effort required vs. blindly polling everything all the time. As a simple example, if a router is down at a node level, you don’t need to waste time trying to check its temperature sensor component status; customizing NerveCenter to do this is very simple and straightforward.