The information design tool in SAP BusinessObjects BI 4.0, the information design tool introduced a new semantic layer for customer environments, as well as new opportunities for end-users to access data and create reports. The 4.2 release has added even more features and enhancements. What are best practices for migrating existing universes onto this tool and getting started with using it? What features are available only in the latest release?
In this Q&A discussion Reporting & Analytics Interactive 2016 speakers Michael Welter and Alan Mayer discussed the features of the information design tool in the latest SAP BusinessObjects BI 4.2 release. Questions included:
Meet the panelists:
- If I create a universe on top of BEx, will it impact performance?
- What is the difference between IDT and UDT?
- Can we connect SP7 server with SP9 client tools?
Michael Welter, Claraview
Michael Welter joined Claraview, a division of Teradata, in 2011, as a Senior PS Consultant. He currently has over 10 years’ experience as a BusinessObjects consultant. He has direct and extensive experience with all facets of the BusinessObjects platform and associated toolsets in all versions from 5.x through the current standard BO 4.x. Michael currently holds multiple BusinessObjects Professional certifications (BOCP). He is also a Certified Instructor and has regularly taught classes in BusinessObjects Administration, Web Intelligence, and other tools.
Alan Mayer, Solid Ground Technologies
Alan Mayer is the founder of Solid Ground Technologies, a firm that specializes in SAP BusinessObjects-based solutions across a variety of industries and platforms. Prior to Solid Ground, Alan was the co-founder of Integra Solutions, one of the first partners recognized by BusinessObjects. He has over 20 years of experience in the BI arena, and participates as the senior architect and principal consultant on some of the largest implementations in the world.
SAPinsiderMatt: Welcome to today’s Q&A on the information design tool in SAP BusinessObjects BI 4.2. I’m pleased to be joined today by Reporting & Analytics Interactive 2016 speakers Michael Welter and Alan Mayer. Please give them a warm welcome!
Please enter your questions into the module below. There are already a number of questions posted on migrating your universes, the settings that will make the tool more user- friendly, and how get the most from the many new features available. We are looking forward to a great event!
Comment From Mileidys SC: How do you assign keys in the universe design when using linked universes?
Michael Welter: Create the keys in the core universe. They will then propagate to the derived universe.
Comment From Maruthi: We come across many issues on multisource universe in 4.1, especially when we used the permute function to convert the date format. These issues including being unable to reach the server and getting failures in the APS data services. Is there anything new in 4.2 that could prevent these issues? :)
Alan Mayer: There are no new innovations in 4.2 for federated universes that I know of. Realize that federated (multisource) universes can be a bit complicated because of the Federator engine behind the scenes. It sounds like the APS you have dedicated for federated services may not be sized correctly.
Comment From Changing report universe: Is there a particular tool you would recommend for resetting the universe connection in all the users' Webi reports? We have 100 or more of them.
Michael Welter: I assume you are asking about converting reports from one universe to another. I have tools from APOS and GB & Smith that do this quite well.
Alan Mayer: The tool from APOS is called the Webi Migrator.
Comment From Guest: If I create a universe on top of BEx, will it impact performance?
Alan Mayer: It will allow for better performance by allowing you to selectively consume the BEx query (vs. taking everything available through a BICS connection in Webi). It’s useful for organizing and formatting output columns as objects
Comment From Guest: What is the difference between IDT and UDT?
Michael Welter: The differences are vast, and too many to cover in this forum. However, I will say this: From the perspective of the end user, there is no difference other than seeing a .UNX at the end. For the developers, however, the basic concepts are the same, but the workflows are quite different. I would suggest searching the web with that question for more comprehensive answers.
Comment From Bill P: Isn't Explorer on the way out and being replaced by Lumira?
Michael Welter: Not yet. In fact, I believe the plan is to fold Explorer into Design Studio. But I may be wrong about that.
Alan Mayer: I think Design Studio and Lumira are converging to one tool - Lumira 2.0.
Comment From Ash: Is it possible to use Java microservices as a source? Do you have a high-level idea of how to use and implement them?
Alan Mayer: Great question, Ash. Not sure I know the answer off the top of my head, but we can find out for you.
Comment From Surya: Hi there. Could you please highlight some of the new features in BO 4.2 IDT and the upcoming ones, too?
Michael Welter: I haven't heard of many new features in IDT in 4.2. But I am pretty excited about the new data connector for the CMS database. I haven't tried it yet, but it's been a long time coming.
Comment From SR: Hello - Is there a way to convert/migrate UNV to multisource IDT?
Michael Welter: No. Single source universes cannot be converted to multisource universes.
Comment From Maria: We have been trying to create a multisource universe but we are experiencing issues (IDT freezes, we are unable to query the universe). Does IDT support 5 connections/data sources at a time, and is there a limit to how many it supports? Our client tools are the same version of BO that is on the network - 4.2 SP3. Thanks!
Alan Mayer: Hi Maria. There is no limit that I know of. Multisource universes, though, require special care and feeding that other universes do not. The Federator engine running behind the scenes processes the queries, translating them into a form of ANSI-92 SQL that is compliant with the tool. I've had problems with earlier versions of IDT on the client side when creating Federated queries that were eventually addressed with a re-installation of the tool and a few tweaks to IDT. Let me see if I can quickly locate my notes on those.
Comment From Balaji: Recently we migrated the from BO 4.0 to BO 4.1. Now when we run huge reports in 4.1, we get CORBA error. Do you know what the reason for that is?
Michael Welter: I haven't heard of that. I would suggest opening a ticket with SAP Tech Support.
Comment From Bill P: Isn't Explore on the way out, replaced by Lumira? per attached image.
Alan Mayer: Hi Bill. That is an old slide - much has changed since. Explorer is on the way out, and current versions of Lumira / Design Studio are converging to Lumira 2.0.
Comment From Balaji: What should be the right version of Java for BO4.2 and BO 4.1?
Michael Welter: Check the PAM for each version. It can change from SP to SP, so make sure you check the correct one for your patch level.
Comment From Surya: I heard a lot about linked universes in recent versions. Is there a good rule of thumb for when to go in for linked universes, and are there any downsides to it?
Alan Mayer: Hi Surya. Linked universes as redesigned by SAP have a lot of advantages, the best of which is reuse of existing business terms and tables/joins. The downsides we faced in the past are mostly taken care of in this 4.2 release. One thing that is not addressed, though, is the replacement of original objects with core objects in your newly derived universe.
Comment From Tim Cummings: We have a universe that is developed by our product team. The Universe is a generic model for our standard database product that we have installed for several clients. Each client has added customizations to our standard database (additional tables, columns, calculation differences). My question is, how can I compare the standard universe I receive from our product team against the client's updated version to determine the differences? Our product team is constantly adding new features to our standard universe, but since I have altered the original universe to suit each client, I cannot deploy the new features unless I manually compare each object. Is there a tool to compare two universes? Is there a way to automatically apply the changes from our product team to my customized client versions?
Michael Welter: There are numerous tools for comparing two universes. I know of at least one that is available on BOB (http://www.forumtopics.com/busobj/index.php). You may also be able to use Visual Difference, a tool available in the CMC. I haven't tried it, but its purpose is to compare different versions of the same object.
As for propagating your changes, I don't know of any simple way to do that.
Comment From Mileidys SC: Again regarding key awareness - I set the primary key in the parent universe, but the fact table to set the foreign key is in the child (linked) universe. Once the universe is linked, every property is disabled and I'm not able to add anything else in the objects
Michael Welter: Yes, that is true and a good point. I don't know of any way around that other than to include the parent universe in the child universe. It's not a great solution, but it may be the only solution at this time.
Comment From Sri: I have a pretty big universe on hand, with about 160 tables, nearly 50 contexts, nearly 360 joins and 7000 objects. Would you recommend converting that or creating a new one?
Alan Mayer: Hi Sri. How easy is it for users to navigate and use that universe? Are they confused because the same information appears multiple times in different classes? The universe becomes their interface to the data sources behind it so their concerns should be listened to. With that said, your universes appears larger than most. That many objects, tables, and joins will take some effort to maintain. I know there are some reasons for keeping universes at that size - ability to create SET-based queries (UNION / INTESECT / MINUS) being one of the hot ones.
Comment From Dan H: is there a release date for Lumira 2.0?
Michael Welter: SAP never announces release dates in advance, just in case the date needs to move.
Comment From Praveen Guntuka: Can we connect SP7 server with SP9 client tools?
Michael Welter: I have done that many times, and it usually works fine. The only time it hasn't worked is when I try to connect 4.1 tools to a 4.2 server.
Comment From tmarkart: What is new in this new version? Or is there somewhere I can see what has changed?
Michael Welter: The SAP Support site, under documentation, has a section on "What's New". You can download a document for each version and SP.
Comment From Mileidys SC: Is there any feature in CMC or an external tool that tells you how many reports are using a defined universe?
Alan Mayer: There are a few methods you could try. The easiest would be to use QueryBuilder to find all the documents based on the universe. Use the CMC to find the ID for that universe. Then run a query like the following in QueryBuilder:
Select * from ci_appobjects, ci_systemobjects where si_id =
One of the columns that comes back will list every document with that universe.
Comment From Mileidys SC: A question about linked universes: Is there any way to change the orders of folders from different linked universes in the derived universe?
Michael Welter: I wish there was, but no, they always go at the bottom.
Alan Mayer: Hi Mileidys. I believe there is a universe parameter that allows you to adjust whether you see the core (inherited) folders first or last.
Comment From Tom: If I convert a UNV to a UNX, what happens to existing scheduled Webi reports? Do we need to manually change each report to run against the UNX universe?
Michael Welter: Keep in mind that when you convert a UNV to a UNX, you are actually making a UNX copy of the UNV universe. So your Webi reports will still be connected to the original UNV. There are tools available from 3rd parties to automate the conversion from one universe to another, as long as the universes are identical. APOS makes one, and GB&Smith makes one. There may be others as well.
Comment From Balaji: Right now we have Webi reports built out of BEx queries. Can I covert the BEx queries to universes and build reports out of them? Will that impact performance?
Alan Mayer: You can create universes on top of the BEx queries already created. It should improve performance from the standpoint that you can selectively consume the BEx query based on the objects selected in the universe query. You can also limit the hierarchy depth within the universe as far as levels retrieved.
Navigation paths, lists of values, and parameters are not supported at this time. Plus, there will be other BEx features that are not available via a universe.
Comment From Mileidys SC: How do linked universes work when I include the core universe? Am I able to sync and get the latest changes? If so, what will happen with the changes I made?
Michael Welter: When you include the core universe, it breaks the link and simply imports the core into the derived. So changes to the core universe will no longer be reflected in the derived universe. That's why I said it's not a great solution. :-(
Comment From Mileidys SC: Is there any way to know what objects from a defined universe are being used in documents?
Alan Mayer: That is a tougher question. It involves opening up and scanning the documents to see which objects were used. This goes beyond the traditional BusinessObjects tool set. SAP offers something called Information Steward that does this for you. My history with that tool is a bit hit and miss. Several third-party tools do a great job. One of the best I've seen for speed and accuracy is MetaMiner from Infolytik.
Comment From Surya: I have already added contexts in IDT but just wanted to make sure of the best practice for adding/dropping joins out of a defined context. I see the signs “+,” “-: and “undefined,” or a blank. Please shed some light on each of these?
Michael Welter: The + sign means that the join is explicitly included in the context, so it will always be a member of that context. The - sign means that the join will be explicitly excluded from the context, and will never be used as part of that context. The blank joins, however, are available to a context, but are not explicitly included. As a rule of thumb, only the joins that directly participate in a loop need to be included in a context.
Comment From Hagit: Do you have a good working demo for the new sets building functionality?
Alan Mayer: Hi Hagit. I will in a few weeks, since that's one of the features I'm highlighting for SAPinsider's Analytics and Reporting 2016 INTERACTIVE conference.
Comment From Balaji: Is there a way to pass a value from one BEx query to another Bex query while running the Webi Report?
Alan Mayer: Not that I know of, Balaji, due to the current restrictions on parameters. One technique we use frequently for other universes is Query on Query, which works fairly well.
Comment From Garry Andrews: We are still using Universe Design Tool primarily for our Universe development. We have developed component universes that we link in our reporting universes. All of the universe developers are running Universe Design Tool on a Citrix server, not on their individual PCs. Quite regularly, 5 times in the last 3 months, we have errors where the reporting universes are corrupted and have to be restored from backup.
Michael Welter: I've not seen universe development done on a Citrix server. Is there a reason for that? I would suggest that each universe have a single owner who is responsible for all development on their universes. That way, they won't step on each other’s toes.
Comment From Surya: I heard some rumblings about BO 4.2 coming pre-packaged with an auditing universe and reports. I do not see them, but I just wanted to confirm.
Michael Welter: I've installed 4.2, and, like you, did not get any audit universe or reports as part of the package. Maybe that will come with a later SP.
Comment From Mileidys SC: Is there any plan for SAP to have a control version system for IDT?
Alan Mayer: Hi Mileidys. One way this can be done today is via Promotion Management. You can use this feature of the CMC to also source-control your universes and documents. It’s not as easy as being integrated with the IDT, but at least it's a start.
Comment From Bernie: Is it possible to have both 4.1 client tools and 4.2 client tools installed on the same computer? If yes, how?
Michael Welter: Yes. They each go into different install folders and different start menu groups. So you can have access to both. I have a different version installed on my laptop, but I sometimes get confused as to which version is currently running. :-)
Comment From Can you tell me if…:I tried the new CMS database for 4.2, but there is no data and not many fields I can use at all. Is it just a placeholder for a CMS database? I am in 4.2 SP3.
Michael Welter: I'm looking forward to trying it. Have you downloaded the sample universe and reports from SAP?
Comment From Balaji: Alan, what do you mean by “Query on Query”? Could you please explain that in a little detail?
Michael Welter: You can use one query as an input to another query. It's similar to a subquery but can be from different sources. You build one query in Web Intelligence, and then, when building the second query, you can drag an object into the Filters Pane and then choose “Equal To” and “Object” from another query.
Comment From Maria: Alan, thank you for your response. If you could provide any notes on how to tweak IDT so that we will not have issues with the multi-source universes, that would be very helpful.
Alan Mayer: I’ve reviewed our past incidents for this problem. We had three problems, all related to connection-level issues with IDT. It seems that IDT connections seem to work best with JDBC. We got into trouble by using ODBC-connections only, then mixing ODBC with JDBC connections.
Comment From Mileidys SC: Thanks Alan. I was looking more for something like a developer tool to see the changes in the universe.
Alan Mayer: Ok. There are still a few options for you. You can use one of the metadata scanning tools mentioned (Information Steward, Metaminer from Infolytik. GB&Smith and EV Technologies also have metadata scanners). The idea here would be to scan different versions of your universe and then perform a DIFF on the results.
Comment From Balaji: Previously in a universe, a single dimension field would allow the user to search for both for text and key-in prompt. But now in a BEx query, I need to have two separate dimensions for the text and key prompts. Is there way to have one dimension for searching both key and text in the Webi prompt? If it’s a BW change, what change should be done for that?
Alan Mayer: Hi Balji. Sorry for answering this last. I don't know of a way off the top of my head, but that doesn't mean that it can't be done.
Comment From Balaji: If the add a combined query feature and the object from another query is grayed out, does it require any BO admin work?
Michael Welter: Combined queries must be enabled in the universe. It is enabled by default, but may have been disabled by the universe developer.
As for query on query, that is available for most, but not all, data sources. I don't believe the administrator needs to do anything to enable it.
Comment From SR: We are facing performance issues using a multi-source universe. We tried to create a simple query with couple of objects, and the performance is very bad. Is there any way we can tune this?
Alan Mayer: There are lots of ways you can tune this. Assuming that your environment has been tuned for federated universes, one of the first places I would look is the DFAT tool (Data Federator Admin Tool). Among other things, this tool can give you the ability to see that the query is being processed and run by the Federator. A query plan of sorts.
SAPinsiderMatt: Thanks to everyone who participated in today’s Q&A! We had a lot of great questions and participation!