Skip to Content

Connecting the on-premise world to the cloud presents many questions and sometimes a picture is the best way to provide an answer. This is why I’ve published architecture diagrams and process flow diagrams for SAP Analytics Cloud consuming SAP BI Platform Universes in a ‘live’ mode.

Live Universe Connectivity Architecture

I’ve presented two diagrams in the wiki:

  • a PATH connection, which will use a Reverse Proxy
  • a DIRECT connection, which uses CORS bypassing the need for a Reverse Proxy

I encourage you to study the diagrams and here are some noteworthy points:

  • The database performance is critical since, the data stays on-premise. It’s never uploaded to the Cloud. Only the model metadata is uploaded to the cloud.
  • The BI Platform uses the WACS (Web Application Container Service) as the gateway into the platform via the REST API that it hosts. This in turn talks to
    • CMS and the FRS (for the obvious reasons)
    • the DSL Bridge Service (to get the universe metadata and map the SAC model to Universe queries)
    • Web Intelligence Processing Server (where a .WID is created and a data provider in that document is created and run against the database for each query asked by the SAP Analytics Cloud client)
  • Most of the load is on the DSL Bridge and the Web Intelligence Processing Server, so you’ll need to make sure you have these services sized appropriately.
  • Each user of SAP Analytics Cloud will connect and login to the BI Platform and thus consume and inherit the security model. Today SAP Analytics Cloud support SAML for SSO, or you can just prompt the user for a user/pass.
  • SSL is mandatory and if your organisations browser policy enforces only trusted certificates, you’ll need to make sure the SSL certificate you use will be valid.


Process Flow diagrams

To really help your understanding of the architecture its best to look at the 5 process flow diagrams I’ve created in the wiki. I walk you through, step by step, from creating a new connection, to creating a model and then using that model to build visualisations and sorting and filtering upon it.

Please do study them. Here’s some additional context to help answer your questions!

  • Actions such as sorting and filtering are performed by the client, the backend isn’t hit.
  • Actions such as filtering of data are all pushed down to the database.
  • The Web Intelligence document is
    • created in the BI Platforms users a favourites folder called “BOELiveProviderTempFolder” (name can be configured).
    • stores the query and results for each query to the database. If you use my Auditing Universe you’ll see a new data provider is created for each query sent. Currently there’s no smart re-use of existing data providers.
    • deleted when the user logs out of SAP Analytics Cloud.

Please do post comments, feedback, questions to this blog rather than one the wiki pages. I will try my best to reply as soon as I can. Please do review other people’s comments and my replies before you post!

As new features and function come into the product I will update the diagrams. I will be publishing a lot more content for consuming Universes, so stay tuned and follow me! 😉

Matthew (Twitter @MattShaw_on_BI)

To report this post you need to login first.


You must be Logged on to comment or reply to a post.

  1. Nick Daniels

    Great information.  My question relates to WebI-in-the-cloud, though.  Is it the same set of functionality as you’d have if fully on premise?  Or is it cut down, I am thinking specifically cut down to the point where traditional 4.1 customers would be unhappy?

    1. Matthew Shaw Post author

      Hello Nick,

      Thanks for your feedback and post. Its always good to hear if the info I share is useful (or not!)

      I’ve not fully understood your question, I think you may need to clarify.

      Today, the Live Connector within SAP Analytics Cloud connects to an on-premise BI Platform Universe. Today, it doesn’t connect to an existing Web Intelligence to consume anything from it. So whilst the live connector today uses Web Intelligence to get data from the database it doesn’t use Web Intelligence as a data source. The data source is the database, via the Universe and via a temporary created Web Intelligence document (that then gets deleted)

      The connector today requires BI 4.2 SP4 or greater (we recommend BI 4.2 SP5 for performance gains). The reason is because the REST API was updated to allow for it to work and so it won’t work with an earlier version.

      Does this answer your question? Regards, Matthew


      1. Nick Daniels


        Heh, my apologies.  What I was trying to get as was this.  If SAP Analytics Cloud is – in simple terms – allowing you to run WebI in the cloud, as opposed to it running on-premise, does it ‘feel’ the same to an end user if they were used to WebI 4.2 on-premise?  I feel out of my depth 🙂

        1. Matthew Shaw Post author

          Hello Nick

          I wouldn’t say SAP Analytics Cloud is allowing you to run Web Intelligence in cloud, at least not yet.

          SAP Analytics Cloud is allowing you to use your on-premise BI Platform to connect and consume Universes and it does so by connecting in a ‘live’ way. Web Intelligence Processing Server is used as the proxy to the database. For an end-user it doens’t look or feel like Web Intelligence, it looks and feels like native SAP Analytics Cloud, it just the data happens is coming from on-premise via a Universe and via Web Intelligence.

          I appropriate it may not be obvious. As I publish more content perhaps it will make more sense. Stay tuned to my blogs and wiki content!

          Regards, Matthew


Leave a Reply