Uploaded image for project: 'Architecture for Learning Enabled Correlation (ALEC)'
  1. Architecture for Learning Enabled Correlation (ALEC)
  2. ALEC-115

Define a list of items to collect from the user. (Alarms / Nodes)

    XMLWordPrintable

Details

    • Story
    • Status: Closed (View Workflow)
    • Minor
    • Resolution: Fixed
    • None
    • None
    • None

    Description

      This list if a fist draft of data that we want to collect to further improve ALEC. Elements that have an asterisk are optional.

      The data would be stored once we've obtained feedback that a situation is correct. Then, we collect the information: Alec setup, each of its alarms, inventory objects and graphs in the situation.

       

      We want data regarding the ALEC setup:

      • Engine (DBSCAN/DeepLearning)
      • OpenNMS instance ID
      • DistanceMeausre (* only if DBSCAN engine)
      • AgreementVersion
      • Epsilon (double)
      • Alpha (double)
      • Beta(double)

      We also want to collect data from alarms.

      • AlarmID (integer)
      • Alarm FirstTimeSeen (timestamp)
      • Alarm LastTimeSeen (timestamp)
      • Severity (string)
      • AlarmSummary (string)
      • AlarmDescription (string)
      • AssignedSituationID (integer)
      • CorrectedSituationID (* only if provided)
      • InventoryObjectID
      • Graph (query GraphAPI with InventoryObjecyID): 3 hops

      Then, we'd like to have some information about the inventory objects in the network:

      • IO Type
      • IO Id
      • IO Subtype
      • IO ParentType
      • List of PeerRefs:
        • Type
        • ID
        • Endpoint
      • List of RelativeRefs:
        • Type
        • ID

      NOTES:

      This issue required no code. This list can change as we identify or disregard data points.

      • Make sure we clean data and protect user privacy. Ex: Remove IP, Ports, etc, node names, etc, etc.

      When it is complete:  / Done Criteria:

      We review the list in the group.  We agree it protects the IP.  We agree it is the most likely to provide us enough information to feed a Machine Learning Engine and run experiments that will further enhance our knowledge of what could be a better engine than DBSCAN.

      What is not:

      • Not the definitive list.  We know it will likely change by the time we actually collect the data.
      • The list will cover the data we think will be needed for the development across the next two or three sprints.

      See ALEC-112

      Attachments

        Issue Links

          Activity

            People

              gmantecon Gerardo Mantecon
              joseanes Jose
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: