Registration

Dear SAP Community Member,
In order to fully benefit from what the SAP Community has to offer, please register at:
http://scn.sap.com
Thank you,
The SAP Community team.
Skip to end of metadata
Go to start of metadata

This page presents a practical step-by-step example for using SAP BusinessObjects Universes with SAP Analytics Cloud.

Please use this page in conjunction with the best practices which presents the theory and more explanations which help answer the 'why' questions

Example 1

We shall:

  1. Create a model on the BI Platform Audit Analysis Universe in SAP Analytics Cloud
  2. Create a story with a number of visualisations

Assumes

  1. The BI Platform Audit Analysis Universe has already been installed (for details on how to install the Universe please refer to the blog)
  2. The Live Universe Connector has already been setup and configured
a sample boe.properties file is available to help.
boe.properties file:
The boe.temporaryDocumentFolder holds the name of the folder that is created inside each users Favourites folder when they consume a Live Universe connection.
–Temporary Web Intelligence documents are created this in folder.
–These documents are then auto-deleted when the user next logon to SAP Analytics Cloud and consumes a Live Universe Connection and the document is at least 2 days old

Best Practice
advice:
  • do not use the boe.compatibleUniverseFolder anymore. The property has been depreciated. Please do not use!
  • New settings to enable a new cache:
    –boe.useUnvListCache and boe.useUniverseCache
    –The boe.useUnvListCache parameter hides, if enabled, universes created after your session has been created.
    –The boe.useUniverseCache parameter hides, if enabled, potential changes made to a universe since it has last been imported
    –Set these to true for production use, false whilst in development. 'True' means you will see all changes as there is no cache, 'False' enables the cache so good for performance, but means you need to logoff and logon again to see any changes.
Installation Guide (also provides the official features supported)

Optionally Modify Universe

  • ŸUniverse Alterations:
    • The BI Platform Auditing Analysis Universe is well designed and needs no special alternations (wink)
    • Please refer to the Best Practices for when alterations may be needed

Validate & Optimise

 

  • Either:
    • Logon to BI Launchpad and login as the user you will use later to connect to the BI Platform from within SAP Analytics Cloud.
    • Select Applications-Web Intelligence, to launch Web Intelligence
  • Or
    • Start Web Intelligence Rich Client
    • Select new document

  • Build the query that you plan to use as the basis of the SAP Analytics Cloud model
  • Add BusinessObjects into the Query Panel ‘Selected Data’:
    • From ‘Time’:
      • Year, Month, Day, Hour, Week Day
    • From ‘Characteristics’:
      • Server, User, Object Type, Application Type
    • From ‘Event’:
      • Event Category, Event Type, Top Level Folder, Object, Status
    • From ‘Measures’:
      • Event Duration (seconds), Document Size (K Bytes), Sessions (Concurrent User), Sessions Peak (Named User + Concurrent), Actions, Rows, Events
  • Add Business Layer filters:
    • From ‘Time’:
      • Only events of the last 12 months
    • From ‘Characteristics’:
      • Only User Activity
  • Run the query

  • ŸAdd a new report (tab) by selecting 'Page Setup' - 'Add Report’

  • In the new report, add the 'Query Summary', by selecting 'Report Elements' - 'Cell' - 'Query Summary' and then adding that into the report by clicking on it
  • Validate the query duration and number of rows returned
  • The query duration and data volume shows good performance so no need to further tune the query or universe

  • Press the Save icon
  • Select the 'Favorites Folder', give the document a name and press Save
  • Delete the document after you have saved it
    • If you are unable to successfully save the document then SAP Analytics Cloud will generate errors to the user

 

Create Models in SAP Analytics Cloud

Create Connection in SAP Analytics Cloud

Login to SAP Analytics Cloud, but very importantly:

  • If you are using a reverse proxy (which will require a 'PATH' connection) then you must login via the Reverse Proxy.
  • If you are using CORS (which will require a 'DIRECT' connection) then you can login directly to the cloud application.
  • Be sure to login with a user that has administration rights to create connections and models.

  • Select Connections from the menu

  • Add a connection by selecting '+' - 'Live Data Connection' - 'SAP Universe'

  • Give the connection a name and description. Importantly note, the name cannot be changed once created.
  • Select the Connection Type, either 'Path' (for use with a Reverse Proxy), or 'Direct' when using CORS

  • If using a Reverse Proxy, the name of the Path Prefix must match that of the path statement as defined in the Apache httpd.conf file.
    • The application will add a '/' to the path if one is not entered. So in this example 'myBIP' and '/myBIP' are both valid and have the same effective value.
  • Select Authentication Method 'User Name and Password' and enter the corresponding details. The username and password are not stored.
    • We recommend using SAML Single Sign On so the user is not prompted for a login to the BI Platform. However for ease of setup we shall use 'user name and password'
  • One language per connection.
    –The default language you select here will not be dynamic, it will be fixed. @Variable('PREFERRED_VIEWING_LOCALE’) will not change.
    –It means all models created with this connection will content to the BI Platform with the locale you select here.
  • Press OK, to save the connection

Create Model in SAP Analytics Cloud

  • Select 'Create' - 'Model''

  • Select 'Use a datasource' and then 'Live Data connection'

  • Select the ‘System Type’ ‘SAP Universe’
  • Select the Connection you created earlier
    • You may be prompted for a User Name and Password to login to the BI Platform if you've not already logged on in this session
  • Select the Data Source as the Universe 'BI Platform Audit Analysis' (if not installed please visit the blog)
    • Please note, a derived universe is no longer necessary.
  • Give the model a name and description
    • The model name can not have any spaces or special characters.
    • Plan to have a good naming strategy to help determine which model is which.
  • Press OK
Universe metadata is uploaded to the Cloud via the browser. There is no direct connection between SAP Analytics Cloud and the BI Platform, it all goes via the browser.
A query panel is presented which will be the basis of the model
Limitations
–The query panel can not be re-sized
–Refer to the query panel feature support matrix for more details

  • Add BusinessObjects into the Query Panel ‘Selected Data’:
    • –From ‘Time’:
      • Year, Month, Day, Hour, Week Day
    • From ‘Characteristics’:
      • Server, User, Object Type, Application Type
    • From ‘Event’:
      • Event Category, Event Type, Top Level Folder, Object, Status
    • From ‘Measures’:
      • Event Duration (seconds), Document Size (K Bytes), Sessions (Concurrent User), Sessions Peak (Named User + Concurrent), Actions, Rows, Events
  • Add Business Layer filters:
    • From ‘Time’:
      • Only events of the last 12 months (Please note: you may need to chose a filter that returns less rows, 12 months could be too much!)
    • From ‘Characteristics’:
      • Only User Activity

  • Press OK
At this moment the query is run against the database and that query is held in a Web Intelligence document (stored in the users favourites folder). That Web Intelligence document becomes a cache for SAP Analytics Cloud. It means that any visualisations built in SAP Analytics Cloud will use that cache rather than querying the underlying database every time a visualisation is changed/amended or created. To force a refresh against the database just refresh the story.
The model is shown.
If you need to go back to the query select the ‘Open Query Editor’ (any changes made to the universe, once republished to the Repository, will be shown after you've logged off and logged back in again)
Select 'All dimension' (we shall modify the dimension to place in them into 'groups')
  • Remove the folder name ('Time' in this example') from the 'Description' of each object, and add it to the 'Group' column
    • You will notice a folder structure appearing on the right hand side
  • Repeat this for all the objects
    • Measures cannot currently be grouped or re-ordered

  • Once all the objects have been ‘Grouped’ the result will be as shown

  • Save the model

Create Story in SAP Analytics Cloud

  • Create a new story

  • Select ‘Access & Explore Data

  • Select the option ‘Data acquired from an existing model’
  • ŸSelect the model you saved earlier

  • Select at least one measure and add dimensions as you see fit
  • Use the Explorer interface to filter, sort, rank, change chart types, add/remove measures and to look at the data by different dimensions
  • Copy the visualisation to a Story Page

  • Within a few minutes you can build yourself a dashboard with an input control to filter the charts as you please

Example 2

We shall:

  • Create two models, each using the same prompt
    • One model will report on ‘Interactive’ sessions, whilst the other will report on ‘Scheduled’ sessions
  • Create a story with a number of visualisations
  • Link the models so the user is prompted just once
  • Define Restricted Measure “Web Intelligence Events” (events where the ‘Object Type’ is ‘Web Intelligence’)
  • Define Calculated Measure “Events per session” (Events / Sessions Peak (Named User+ Concurrent)”

Assumes

  • Completed Example 1

Copy existing model twice and edit new models (Interactive and Scheduled)

  • Browse the list of models

  • Search for the model created in example 1
  • Select the model (Live_BI_Platform_AuditAnalysis)
  • Select copy the model (you will need to repeat this step in just a moment)

  • Give the new model a name 'AuditAnalysis_Interactive'
  • Copy the model again and give the new model a name 'AuditAnalysis_Scheduled'

At this point you will have 3 identical models, except they will have different names. The 'query panel' definition has been copied with the model.

  • Browse the models
  • Open the model you just created called 'AuditAnalysis_Interactive' (we need to edit the filters so this model only returns Interactive sessions)

  • Open the query editor
  • In the filters pane of the query panel remove the filter called 'Only events of the last 12 months'

  • Add the filter object 'Interactive Mode (Only Interactive Sessions not Scheduled Sessions)'
    • This object has a Universe Prompt defined in it
  • Press OK

  • Save the model

  • Browse the models
  • Open the model you created called 'AuditAnalysis_Scheduled' (we need to edit the filters so this model only returns Scheduled sessions)

  • Open the query editor
  • In the filters pane of the query panel remove the filter called 'Only events of the last 12 months'

  • Add the filter object 'Interactive Mode (Only Interactive Sessions not Scheduled Sessions)'
    • This object has a Universe Prompt defined in it
  • Press OK

  • Save the model

Create new story

  • Create a new Story

  • Select ‘Access & Explore Data

  • Select 'Chart'

  • Select the model 'AuditAnalysis_Interactive' (this returns on Interactive sessions and press OK

  • Select the option 'Automatically open prompt when story opens'
  • Answer the prompts by using the calendar pop-up
  • Press Set

  • Select 'Combination Column & Line' Chart type
  • For the Measure 'Column Axis' select 'Measure/Events'
  • For 'Dimensions' select 'Time/Hour'
  • Swap the colours (so to make the chart more attractive)

  • The chart should look like this.
  • Position it to the left hand side of the canvas page

  • Insert a new 'Chart'
  • Change the chart type to 'Combination Stacked Column & Line'
  • Select the icon for 'Change Primary Model'
  • Select the 'AuditAnalysis_Scheduled' model and press OK

  • Select the option 'Automatically open prompt when story opens'
  • Answer the prompts by using the calendar pop-up
  • Press Set

  • For the Measure 'Column Axis' select 'Measure/Events'
  • For 'Dimensions' select 'Time/Hour'
  • Swap the colours (so to make the chart more attractive)

  • Position the latest chart to the right hand side of the chart created earlier

Link Variables across both models

  • Select 'Edit-Prompts'-Link Variables...'
    • Currently the story has 2 models in it. Each model has 1 set of prompts (with 2 prompts in each set) and each model will prompt the user giving the user 2 sets of prompts to answer.
    • However the prompts are the same, so if we link them the user will only need to answer one set of prompts, not two.

  • Select 'AuditAnalysis_Interactive' as one of the models, and 'AuditAnalysis_Scheduled' as the other model.
  • Select 'Event Started After' and 'Event Started After' (from each model)
  • Select 'Event Started Before' and 'Event Started Before' (from each model)
    • (the order you select is important. it means you need to select each pair before moving to the next pair of prompts)
  • Press Set

  • Set the linked variables
  • Press Set
  • Press Done

Create new Restricted Measure

  • With the chart on the left hand side (Interactive sessions) open the Builder and under measures select 'Create Calculation'

  • Select the type 'Restricted Measure'

  • Give the measure a name 'Web Intelligence Events'
  • Under 'Measure' select 'Measures/Events'
  • Under 'Dimensions' select 'Characteristics/Object Type'
  • Then select 'the value 'Web Intelligence'
  • Press OK

This restricted measure will show the number of events for the Object Type 'Web Intelligence'

The Story will look something like this

Create new Calculated Measure

  • Under the 'Line Axis', select 'Create Calculation...'

  • Select the type 'Calculated Measure'
  • Give the measure a name 'Events per session'
  • In the Edit Formula box, press CTRL + SPACE to pop-up the list of measures
  • Select 'Measures/Events'

  • Enter a /
  • press CTRL + SPACE to pop-up the list of measures again
  • Select 'Measures/Sessions Peak (Named User+ Concurrent)'

  • Press OK

Add Text Box with Dynamic Text of Prompt Values

  • Add a 'Text' box

  • Add the text 'Interactive and Scheduled Events Between'
  • Change the font size, alignment and colour as you please
  • With the cursor at the end off the text 'Interactive and Scheduled Events Between' (i.e. where you want the dynamic text to appear), Select the '...' and 'Dynamic Text'

  • Select the Variables 'Events Started After' and 'Events Started Before'
  • Press Create

Create Measure Input Control

  • With the chart on the right hand side (Scheduled Events), under 'Line Axis' and under 'Measures' select 'Create Measure Input Control'

  • Select the measures 'Actions', 'Document Size (K Bytes)', 'Event Duration (seconds)' and 'Rows'
    • (the other measures are not relevant for scheduled events and 'events' is already shown as a bar chart)

The end result!

This wiki is referenced by a blog. Typically please comment on the blog, not directly on these wiki pages.

  • No labels