Events & Paging Data in Analytics - Understanding the Invalid Joins
This document assumes you are familiar with Shoplogix Event data, and/or Paging/Andon data.
Events Data are now live (Release date: April 16, 2021) in Analytics. Paging released in May 2021.
Due to the many-to-many nature of the primary [Fact Core] to [Fact Events] relationship, many uses of the Fact Events table are not valid with our [Fact Core]-centric model. Same join restrictions for Paging.
The following joins are valid:
[Dim Machines]
[Dim Dates] (use [Fact Events].[Event Start DateTime] / [Dim Pages].[Page Created On])
the following joins are INVALID:
[Dim Jobs]
[Dim Shift Instances]
[Dim Recent Shifts]
Attempting the invalid joins will duplicate results and give confusing results. The supported use cases for Events are:
Event Pareto: All occurrences, durations, and average duration
Filterable by Machine Name and Event Start DateTime.
Note any correlation to Shift or Job is best achieved by identifying time ranges and filtering to those. Shifts and Jobs not at the hourly level will not be able to perfectly correlate.
Totally reasonable thing to ask for, but not currently possible, is “show me all the event durations for these 3 shifts.” We may be able to build enhancement to support this. CS-Eng, please add a comment and tag @Steven Arrol (Unlicensed) @Ben Patterson (Unlicensed) to draw attention to your need.
Historical Events and Paging Data
Events are saved for 60 days in analytics.
Pages have full history.
Events can be exported from the exportapi if events older than 60 days are required.
More Info
Engineers, here is the full background in context of Analytics: https://slxdev.atlassian.net/browse/PROD-244