For the list of all supported Decision Insight downloads and releases, see the Downloads page.


Rhythms define the granularity of time within an application and within components of the application. Within Decision Insight’s built-in time dimensions, rhythms set the timing of when data is both analyzed and refreshed.

Rhythms are defined based on the needs of the users and stored in a space. In a way, they can be visualized as a type of sine wave, where the frequency of crossing the axis in a positive direction defines the rhythm. Decision Insight has the inherent ability to manage multiple rhythms within the same application. In many cases, rhythms constitute a hierarchy of cycles.

An application runs at the pace of the smallest rhythm. For most applications, the shortest interval is 5 minutes. Some may be longer; few will be shorter.

Rhythms do not affect information arriving into the solution. Events that arrive at sub-second intervals are still monitored and absorbed – however, the information display in Decision Insight is only refreshed based on the base rhythm. Matching the shortest interval to the pace of the process is important because your application will never analyze or refresh data at a finer granularity. If you pick 15 minutes as the shortest interval and the users expect to see data updates every 5 minutes, they will be disappointed with the implementation. 

This is an area where Axway’s expertise is important. If you ask a user how frequently they want to see data updated on their dashboards they will often pick an unrealistic number. You may hear anything from "once a minute" to "in real time." With very (very, very) few exceptions there is no value in updating information more frequently than 5 minutes. Users will suggest shorter times mostly because what they don’t want is an end-of-day report.

Looking towards the value and purpose of a Decision Insight solution, you know that the goal is to provide actionable information. The actions that can be triggered by a change in a Decision Insight dashboard occur on a human scale. In most business circumstances, five minutes falls well within expected reaction times. In those rare situations where a maximum of five minutes before notification could impact business outcomes, it is difficult to imagine a situation in which a rhythm smaller than one minute would ever be required. Remember that the goal is for a person to initiate an action in response to a notification or dashboard change; human beings really do not operate at sub-second intervals. Air traffic control systems require rapid updates but it is unlikely you will find an implementation that involves things of that nature.

A good practice is to define rhythms that are logical multiples of shorter intervals. For example, a design could use rhythms where each longer rhythm is twice as long as the previous shorter one. This would create rhythms of 5 minutes, 10 minutes, 20 minutes, 40 minutes, and 80 minutes. This has a certain mathematical purity but no user is likely to understand an interval of 80 minutes as opposed to 1 hour.

The following section illustrates how different duration can be combined in defining rhythms. Imagine that a process been defined with a 5-minute rhythm as the finest granularity. You can think of this five-minute granularity as a series of buckets that each hold information for that time range.

This five-minute definition sets the heartbeat of the application. These time slices will hold information about everything that happened or was detected or evaluated during that interval.

Important: Events are reported at the end of a rhythm interval. This means that a file arriving at 10:00:01 will be visible at 10:05:00. This is because there is no way to predict at the start of interval what will happen during the next five minutes. It is worth remembering when data starts refreshing when you are using (or explaining) the content of a dashboard.

When you configure rhythms with a longer interval, you can think of them as being a summary of all the information available during the finer granularity intervals:

Each of the 15-minute intervals consolidates information from, as shown above, three 5- minute intervals. Four of the 15-minute intervals, in turn, are consolidated into a 1-hour rhythm. A configuration might include rhythms for an hour, a day and a week, defined to match the analysis most useful by the observer.

Related Links