Setting upper or lower threshold seems to be simple, right? What happens when the value is oscillating and goes over and under the threshold boundary every minute? Unfortunately, you get an alert every other minute.
This situation is just an example of why some simple models fall shortly sometimes, and you need to reach for a bit more advanced features like thresholds with separate reset boundary.
Another example might be when you want to react only if both of two connections fail, otherwise you need just a notification. Advanced monitoring algorithms are the answer to these specific needs that might be not frequent but sometimes indispensable.
Business Status Nodeto represent and calculate the status from multiple monitored elements.
NetCrunch expands beyond traditional network device monitoring. Practically it's able to monitor any process and devices where reaction time does need to be shorter than 15 seconds.
NetCrunch can process events from many external sources that can behave in the unusual way. That's why we have equipped NetCrunch in a range of additional conditions that you can add to any alert. NetCrunch provides conditional alerting which allows monitoring heartbeats or react to missing events. It also can make the alert less sensitive to repetitive events.
All NetCrunch internal alerts are correlated. For other, external alerts (i.e. traps, syslog messages) it allows adding a list of actions to the alert that can close the alert. The alert can also be closed automatically after a given time. This kind of events have to occur on the same node.
Another correlation can be defined for alerts happening on different nodes within the same time frame.
Read more at Alert Conditions & Correlation
A node in NetCrunch is an Atlas entity, a group of services usually representing a network endpoint (interface)
Nevertheless, NetCrunch node can represent much more than just a network endpoint now, as the network monitoring model has evolved from simply looking at devices, then endpoints, to finally becoming the monitoring of a cloud of services.
The node often doesn't need an IP address now.
Such kind of node is a
Business Status node that represents a state of a group of services (nodes, sensors service, or monitoring packs) that is calculated automatically.
You can add multiple elements to each group which translate to a logical condition determining the state of the node.
Business status can rely on another status node so you can simply create a path of critical services by building a chain of Business Status nodes.
Unlike entry-level monitoring products, NetCrunch offers a vast range of threshold conditions, including baselines for professional usage.
These thresholds will help you avoid false positive alerts and detect abnormal behavior or abrupt change of the monitored value,
NetCrunch is built upon solid statistical analysis. The program offers high accuracy with somewhat lower precision to achieve the high level of performance.
NetCrunch stores both raw data and pre-computed data samples. Depending on the level of analysis, the program uses different data streams to accurately present data on the screen (you need no more data than your screen can display).