Lumira 2 Workflows

[Matthew_Shaw]

Lumira 2 saw Lumira 1.x and Design Studio 1.x converge into a single technology.

These two different technologies had their own BI Platform service and utilised other services on the BI Platform in a different ways, depending upon the workflow.  With the technologies now converged all these workflows have changed.  So anyone updating to Lumira 2 will be keen to understand what processes are used when and for what reason.

I have created 26 workflow diagrams showing which process is used when. These Lumira 2 workflow diagrams are available in the SAP Community Wiki.

In summary they cover workflows for the client interfaces:

• Lumira Discovery


• Lumira Designer


• BI Launchpad for both Discovery and Designer authored documents

For each of these they detail which process is used when for:

• Live connections (BW, HANA, Single and Multisource Universe)


• Import data connections  (BW, HANA, Single and Multisource Universe, Excel, Query with SQL)

What's worth noting?

• The Adaptive Processing Service 'DSL Bridge' isn't used for any workflow. This is because these libraries are now included internally within the client. This in turn means one less 'hop' and an improved performance.


• The Lumira Processing Server (on the BI Platform) isn't used when connecting, from the client, to either HANA or BW. The connection is direct.


• When saving a document to the BI Platform, the WACS (Web Application Container Service) is used using a REST API. So before you may not have necessarily needed this service, however you will with Lumira 2.


• Lumira Discovery uses CORBA to login, whereas Lumira Designer uses SOAP to the dswsbobje web application. (Over time you should see a convergence to use the REST API for these workflows)


• Connectivity to HANA is using InA protocol for online (no longer JDBC) and JDBC for 'download'


• Connectivity to BW is using BICS protocol for online and JDBC for 'download'

Your feedback is important and I will do my best to reply and keep the diagrams up-to-date and correct any mistakes.

Feel free to post a general question to this blog and post a comment within the page of the particular workflow if you have a particular question about that workflow.

Feel free to 'like' this page if you do!

Matthew Shaw (Twitter: @MattShaw_on_BI)

 

[former_member309335]

Thanks Matt for a beautiful blog on Lumira 2.0. I like the section particularly on "What's worth noting?"

[Matthew_Shaw]

Many thanks Santhosh for your feedback

 

[quovadis]

Thank you Matthew for the blog; I have a small question related to HANA DB JDBC offline connection and the Lumira 2.0 server add-on on the BIP platform (aka the Lumira Processing Server)

When you say "

• The Lumira Processing Server (on the BI Platform) isn’t used when connecting, from the client, to either HANA or BW. The connection is direct. "

a. this applies to both HANA and BWoH/B/4HANA over the INA connection ?

b. HANA JDBC connection will still be established by the Lumira Processing Server when serving the LUMX from the repository

Thx

Piotr

[quovadis]

Hmm. Looking at the  Lumira 2 workflow diagrams are available in the SAP Community Wiki 

it seems it is the other way around. The offline (import) HANA / BW connection over JDBC is established directly bypassing the Lumira Processing Server.

thx

Piotr

[Matthew_Shaw]

Hello Piotr

a. this applies to both HANA and BWoH/B/4HANA over the INA connection ?

Yes.

b. HANA JDBC connection will still be established by the Lumira Processing Server when serving the LUMX from the repository

 I’m a little confused by the question as its slightly ambiguous. I’m not sure what to mean by ‘serving the LUMX’, I presume you mean saving the document to the repository from the desktop client . When the document is saved to the Repository from the desktop client there is no connection to any query database, however the Lumira Server is used as part of the same workflow, although no real work is done by the Lumira server. The Lumira server just needs to be running and enabled to allow the workflow of saving to be successful.

Feel free to comment on the individual workflow within the wiki and mention the step number (1,2,3 etc.) that I put into each of the workflows to reduce the chances of misunderstandings etc.

Regards, Matthew

[former_member435026]

Hey Matthew,

 

Do you know of any resource with information on configuring a JDBC driver for Lumira Server for BIP (where the Lumira Server is on a separate host to the rest of SAP BI).  Ive run through the config manual but the steps in there dont seem to work!  Ive not found anyone doing the same thing either which doesnt help.

[Matthew_Shaw]

Hello Andy, You might well need to install some additional main Platform components to get the 'ConnectionServer' and all its related files installed. I strongly recommend installing everything everywhere. I am aware that a install of just the most basic BI Platform plus the Lumira add-on can lead to issues. Even if you're not using any of the main BI Platform services on your Lumira server, then I still recommend installing the lot! Within the CMC, you can delete all unnecessary services, so you just have the Lumira service running.  If your using a Universe, you can see from the process flows, that the connection server is needed. So it makes sense to run the Connection Server on the same machine as Lumira (you might need it on the other machines too of course)

Regards, Matthew

 

[former_member435026]

Thanks very much Matthew, Ill try this approach, shame it wasnt suggested by the support guys!

[former_member435026]

Hi Matthew (and anyone else who hits the same problem).

 

Installing a connection server(s)  on my Lumira 2.0 Server for BIP has now solved all my connectivity issues (both direct JDBC and UNX).  Thanks very much for your suggestion!

 

Andy

[former_member435026]

Would you believe this was all working yesterday, come to work with the system again today and none of the Lumira Designer, based on Universe documents work anymore, yet Lumira Discovery based on direct JDBC still work ok.  None of which makes sense at all.  Dont seem to be able to consume a Universe properly either, lets me define a query, and even preview but gets an 'Analysis Application' error when trying to use the query within the document.  All seems very variable and unreliable, wouldnt be able to recommend any of this in a production environment deployed in this way, especially without any proper documentation.....

[Matthew_Shaw]

A few ideas if I may

With the process flows you can see which services are used and for what workflow. For you the connection server on the BI Platform will be used. So first test the connection is ok in IDT, then also test the application in Lumira Designer. Somewhere en-route you should find an issue (password, connection server not running/enabled).

Regards Matthew

[martin_dingeldey]

Hi together,

I've the same problem with the distributed install, like it was working with Lumira 1.31. and it is described in the AdminGuide.

I opened a case for it and will inform you all if there is a solution or which components have to be installed not to hava all the unnecessary stuff installed with the platform

@Matt:Thank you for all the Infos you're sharing, it's one of the most worthy source for all Information about BO Updates und Lumira 2.0 Stuff. Thank you very much!

Regards,

Martin

[sachin_v_joshi]

Hi Matthew,

You have presented Lumira 2.0 workflows but has SAP forgot to publish the same workflows for 1.31.6 version.

I could not find in the SAP wiki pages. Any idea where can i get them for HANA, SAP BW ,Excel , Unx data sources ?

Cheers, Sachin

[Matthew_Shaw]

Hello Sachin

I'm sorry to say, SAP haven't published the same workflows for Lumira 1.x, so you'll not find them anywhere. The workflows will be similar but not identical.

Regards, Matthew

 

[former_member341062]

Hi Matthew,

I was looking at the presentations, and I was wondering if you might have something more detailed on the process flow between server components when opening a Lumira Discovery created document through the Lauchpad.

We are trying to upgrade to full Lumira 2.1 from a landscape with Design studio only. The primary BO workload, webserver (NetWeaver) and CMS server are on IBM AIX. AIX has been upgraded to Lumira Server from Design studio and we have added a new Windows node for Lumira Server to provide for Discovery, but are facing issues. The first call to open the Lumira Server webapp or a Discovery created document through the Lauchpad after a server restart will work. After that the response will be "Cannot open Discovery documents with AIX / Solaris". The Discovery client tool can download the same document from BI platform and view it correctly.

So if possible, could you provide an in-depth process flow for a launchpad view call of a discovery doc, or just tell me what the problem is, whichever is the easier 🙂

 

Best regards,

Ola

[Matthew_Shaw]

Hello Ola

I believe, though not 100% sure, the issue you described is a known issue and I think was resolved in either Lumira 2.1 or a subsequent patch/SP. If you check the list of fixed issues that might help clarify that. Log a support incident if you can't find it listed.

I believe the problem you have is related to the fact you have a deployment where the Lumira software is installed on a machine with only some of the BI Platform installed software components or 'features'. I believe that you need to install many more BI Platform software features, for example the Connection Server and perhaps some others. In general, as a I mention in my blog, its always a good idea to install everything everywhere. By this I mean install all the software components for the BI Platform and for the Add-ons on all the machines in the cluster (except for the Web App Server). This gives you the most flexibility and consistency. For any services/servers you don't want to have running on any particular machines, you can just delete them from within the CMC.

Regards, Matthew

[Former Member]

Awesome blog! This is really great info about which server handles what etc. Thanks.