Live Universe connectivity, to the SAP BI Platform, is about to get a significant update but watch out! If you don't take action before wave 14 hits you, you won't be able to edit or create new Live Universe models.
These changes are planned to be available in Wave 14, but importantly if you already connect to a Universe with a Live connection, you may need to take action BEFORE wave 14.
This blog introduces what's new and what you need to do before wave 14.
As mentioned in the introduction there's two main new features, but there's a few more too!
When creating a new Model, off a Live Universe Connection, a query panel is now presented. It means you can
It means you don't need a derived universe anymore. It also means you can query multiple fact tables as contexts are now supported, including context prompting. All that's required is a single result set.
The query panel does have some limitations and I've documented them in my wiki under 'Query as a Data Source Supported Features'
When a user uses the model, the query that defines that model is run by Web Intelligence on the BI Platform. The query result is 'cached' in a Web Intelligence document that is held in the users BI Platform favorites folder. This cache is used every time SAP Analytics Cloud wants to display a visualisation. Before, SAP Analytics Cloud would send the query to the underlying database every time.
Consequently the users experience should be dramatically improved since Web Intelligence will be serving up the data, rather than asking the database every time. The only cost is the user may have to wait longer for initial query to be run.
Restricted measures enables you to create a measure on the condition of a dimension. In this example I've created a 'Web Intelligence Events' measure which only shows the 'Events' when the 'Object Type' is 'Web Intelligence'.
Calculated measures enables you to build new formula on existing measures or other calculated or restricted measures.
The logic for this is pushed down from SAP Analytics Cloud to Web Intelligence. This feature required SAP BI Platform 4.2 Support Pack 4 Patch 3.
With two models containing the same universe prompts it's nice for the end user to only need to answer the same prompt once, rather than once for each model they have in their story.
Thanks to 'Link Variables' you can now link variables so only one set of universe prompts is provided.
Watch out! 'Model Linking' is not yet supported, though the User Interface allows you to link models. This is 'work in progress'! 😉
Stat tuned for the official documentation to come, but I have updated my wiki with the new features in the 'general' section.
Your existing and any new stories you build will continue to work. (That is content created on or before SAP Analytics Cloud wave 14 with the Universe Connector v1.04 but working with an SAP Analytics Cloud wave later than 14)
However if you try to edit any existing Live Universe models or create new Live Universe model you won't be able to do so. You'll have to update your on-premise Connector.
With the existing SAP Analytics Cloud wave (up to and including wave 13) the version of the on-premise Live Universe Connector is v1.04. This is the version of the cs_ina_agent.war (as deployed on Tomcat)
It means you should update the on-premise 'Live Connector' before wave 14 if you want to create new or edit existing Live Universe models.
The latest release is available here in KBA 2492800. Today that is version 1.0.6 (or v1 SP6)
Perhaps confusingly, an SAP Analytics Cloud tenant earlier than wave 14 can use any version of Live Universe Connector. For example, if you updated to version 1.0.6, but you had an older tenant than wave 14, then it is supported, it will work. It's just that it will work as the old version did. You'll only get access to the new functionality described above when the tenant is updated to Wave 14.
This might be a little confusing, so I've create a matrix table of 'Universe Live Connector Support Matrix'
There is a known issue with version 1.0.6. The issue is related only to when
We're still isolating the issue, but if you get a warning saying the 'version is out of date' then try version 1.0.5 until we identify the root cause.
[Update: 5 September 2018]: The next version of the cs_ina_agent.war, version 1.0.7 is now available on SMP which resolves this issue.
When wave 14 hits your tenant all you need to do is edit the model to 'upgrade' it and benefit from all the new features. When you edit the model you'll be presented with the Query Panel to define the model definition.
With all the excitement of wave 14 for Live Universes I thought it would be great for you to see what was coming by means of providing some step-by-step examples. I provide 2 detailed examples:
Feedback is much appreciated, as too are questions.
Matthew
Twitter @MattShaw_on_BI
- A 'query' panel to define what makes up the SAP Analytics Cloud model resolving the 'single context' restriction of past
- A subtle but important change in the architecture resulting in a potentially significant performance boost for end users
These changes are planned to be available in Wave 14, but importantly if you already connect to a Universe with a Live connection, you may need to take action BEFORE wave 14.
This blog introduces what's new and what you need to do before wave 14.
What's planned for Live Universe Connectivity?
As mentioned in the introduction there's two main new features, but there's a few more too!
Query Panel for Modelling
When creating a new Model, off a Live Universe Connection, a query panel is now presented. It means you can
- define the scope of the model by adding whichever objects you want and of course leaving some out! (With the existing version all objects where part of the Model)
- add filters and nest conditions with AND/ORs (With the existing version this wasn't possible at all)
It means you don't need a derived universe anymore. It also means you can query multiple fact tables as contexts are now supported, including context prompting. All that's required is a single result set.
The query panel does have some limitations and I've documented them in my wiki under 'Query as a Data Source Supported Features'
Performance
When a user uses the model, the query that defines that model is run by Web Intelligence on the BI Platform. The query result is 'cached' in a Web Intelligence document that is held in the users BI Platform favorites folder. This cache is used every time SAP Analytics Cloud wants to display a visualisation. Before, SAP Analytics Cloud would send the query to the underlying database every time.
Consequently the users experience should be dramatically improved since Web Intelligence will be serving up the data, rather than asking the database every time. The only cost is the user may have to wait longer for initial query to be run.
Other goodies
Restricted measures
Restricted measures enables you to create a measure on the condition of a dimension. In this example I've created a 'Web Intelligence Events' measure which only shows the 'Events' when the 'Object Type' is 'Web Intelligence'.
Calculated measures
Calculated measures enables you to build new formula on existing measures or other calculated or restricted measures.
The logic for this is pushed down from SAP Analytics Cloud to Web Intelligence. This feature required SAP BI Platform 4.2 Support Pack 4 Patch 3.
Linked Variables
With two models containing the same universe prompts it's nice for the end user to only need to answer the same prompt once, rather than once for each model they have in their story.
Thanks to 'Link Variables' you can now link variables so only one set of universe prompts is provided.
Watch out! 'Model Linking' is not yet supported, though the User Interface allows you to link models. This is 'work in progress'! 😉
More information?
Stat tuned for the official documentation to come, but I have updated my wiki with the new features in the 'general' section.
I'm using Live Universe Connections today will it carry on working?
Your existing and any new stories you build will continue to work. (That is content created on or before SAP Analytics Cloud wave 14 with the Universe Connector v1.04 but working with an SAP Analytics Cloud wave later than 14)
However if you try to edit any existing Live Universe models or create new Live Universe model you won't be able to do so. You'll have to update your on-premise Connector.
With the existing SAP Analytics Cloud wave (up to and including wave 13) the version of the on-premise Live Universe Connector is v1.04. This is the version of the cs_ina_agent.war (as deployed on Tomcat)
It means you should update the on-premise 'Live Connector' before wave 14 if you want to create new or edit existing Live Universe models.
The latest release is available here in KBA 2492800. Today that is version 1.0.6 (or v1 SP6)
Perhaps confusingly, an SAP Analytics Cloud tenant earlier than wave 14 can use any version of Live Universe Connector. For example, if you updated to version 1.0.6, but you had an older tenant than wave 14, then it is supported, it will work. It's just that it will work as the old version did. You'll only get access to the new functionality described above when the tenant is updated to Wave 14.
This might be a little confusing, so I've create a matrix table of 'Universe Live Connector Support Matrix'
Known issue with version 1.0.6?
There is a known issue with version 1.0.6. The issue is related only to when
- cs_ina_agent.war is deployed into the same tomcat as the BI Platform 'BI Launchpad'
- AND (BI Platform 4.2 Support Pack 5 or BI Platform 4.2 Support Pack 6).
We're still isolating the issue, but if you get a warning saying the 'version is out of date' then try version 1.0.5 until we identify the root cause.
[Update: 5 September 2018]: The next version of the cs_ina_agent.war, version 1.0.7 is now available on SMP which resolves this issue.
How do I 'upgrade' my Models?
When wave 14 hits your tenant all you need to do is edit the model to 'upgrade' it and benefit from all the new features. When you edit the model you'll be presented with the Query Panel to define the model definition.
Detailed Step-by Step examples
With all the excitement of wave 14 for Live Universes I thought it would be great for you to see what was coming by means of providing some step-by-step examples. I provide 2 detailed examples:
- Creating a new model and story
- Copying a model twice, modifying the model query, defining linked variables, creating restricted and calculated measures
Feedback
Feedback is much appreciated, as too are questions.
Matthew
Twitter @MattShaw_on_BI
- SAP Managed Tags:
12 Comments
