ALEC 2.1-alpha Invitation

ALEC 2.1-alpha is available and we want to get your feedback in order to make a better release. All feedback is welcome: you can try it on your own, or I can give you a demo on my system, or both. Please contact me directly through discourse or @jose at Additionally, ALEC is available in RPM form or Debian packages from the repositories.

Release 2.1 Description and Notes
ALEC 2.1 improves the visualization of situations and the alarms affecting them by:

  • New GUI - We have re-designed the way we show situations under OpenNMS Horizon new graphical User Interface. This will help you identify active situations, the complexity and severity of them, and how alarms relate to each other.
  • Situation Timeline - We help you visualize at what time the alarms that contribute to a situation started, supporting your efforts at determining the root cause of a situation.
  • Accept or Reject a Situation - You now have the opportunity to Accept a situation, and most importantly Reject the situation. Upon Rejecting it you won’t be bothered by it anymore, and it will give you an opportunity to help the OpenNMS product (more on this below).

In addition, it incrementally improves on the DBSCAN alarm correlation engine by allowing you to select the Hellinger Distance as a mechanism to detect what alarms may be related to each other.

Most importantly, with your voluntary help, we will be able to capture information about the situations that you believe are correct and the ones that we have mis-identified. This will help us bring you a future version of the product that is more accurate at detecting situations.

ALEC - Architecture for Learning Enabled Correlation - A platform where we use Machine Learning to tackle the Sea of Red caused by a large number of alarms and correlate them into Actionable Insight, through Situations.

Situation - A situation is a group of alarms that we suspect are related to each other and are a consequence of the same root cause. By grouping them together we support you in quickly identifying the source of the problem, addressing them, and as such reducing the complexities and noise resulting from a large number of alarms.

1 Like