Cognos Framework Manager vs. Data Modules
Can You Use Data Modules as the Primary Cognos Metadata Modeling Tool?
Each release of Cognos sees incremental improvements to data modules. Over time, the functionality gaps between data modules and Framework Manager have narrowed substantially. In fact, data modules are a powerhouse of a tool.
But does that mean you can use data modules as the primary Cognos metadata modeling tool? Now is your chance to find out.
In this on-demand demo and comprehensive comparison between Framework Manager and data modules, we explore data module advantages over Framework Manager and the feature and functionality gaps that exist. And answer participant questions.
Cognos Framework Manager
Cognos data modules
Senior BI Architect
Pedro joined Senturus in 2010 and brings over 20 years of BI and data warehousing experience to his role. He has been instrumental in implementing data warehousing systems from scratch and has experienced the evolution of the BI industry through several iterations of BI products including Cognos, MicroStrategy and Tableau.
Greetings everyone and welcome to this latest installment of the Senturus knowledge series. Today, we’re excited to be presenting to you on the topic of Cognos Framework Manager versus data modules. We’ll do a comprehensive comparison and discuss some key feature gaps.
Before we get into the presentation, a few housekeeping items, please feel free to use the GoToWebinar control panel to make this session interactive. We’re usually able to respond to your questions while the webinar is in progress, and encourage you to enter your questions via the question pane in that Control Panel. Which, you can minimize or restore using the orange arrow.
If however, we’re unable to respond to your question during the live webinar, we will cover it in a written response document that we post on senturus.com.
Which brings us to the next question we get early and often throughout the presentation, is, can I get a copy of the presentation? And the answer is absolutely.
It will be available shortly on senturus.com. You can select the resources tab and head to the resources library. Or you can click on the link that we’ll put in the GoToWebinar control panel if we haven’t already and be sure while you’re there to bookmark the resource library as it has tons of valuable content addressing a wide variety of business analytics topics.
Our agenda, today, after some quick introductions, we’ll do an introduction of those Framework Manager and data modules and provide some comparisons between the two.
Discuss some key framework manager model issues versus data modules, and demonstrate some solutions to give you an idea of what that looks like in real time. And then stick around for the Senturus overview, for those of you who may not be familiar with who we are, and what we do. Some additional, almost entirely free resources, and then at the end, we always have our great Q and A so, again, get your questions in the Q and A, and we’ll get to those at the end of the presentation.
So joining me today, I’m pleased to be joined by mister Pedro Ining , Pedro joins Senturus back in 2010 and brings over 20 years of BI and data warehousing experience to his role. In addition to being a regular contributor to our Senturus is Knowledge Series here. He’s been instrumental in implementing data warehousing system from scratch and has experienced the evolution of the BI industry through several iterations of BI products, including Cognos, MicroStrategy, and Tableau.
My name is Mike Weinhauer. I wear a number of different hats at Senturus. One of them being the emcee for the Knowledge Series, and I’m pleased to be here hosting you today. And with that, I’ll hand the floor and the mic over to Pedro.
Pedro floor is yours OK, hi, everybody. So today, Framework Manager versus data modules.
It’s worth mentioning that Cognos has been around for quite awhile since the mid nineties and has quite a bit of a modeling legacy as products that have been around for a while and have morphed with the industry contain.
So, we all remember modeling with transformer and then Framework Manager and with DMR, we’re just trying to make Framework Manager look like transformer who can never forget dynamic cubes and anybody using that out there. So, dynamic cubes hasn’t been used for awhile.
We have Cognos data modules with the Cognos Analytics 11.
The product has evolved quite a bit over the years and we always have to change our ways and how to work with these tools.
IBM has really pushed data Models.
It doesn’t really, really advertise Framework Manager as part of the 11, as a feature set, but it’s there.
And we’re kind of taking the point of, you know, if you’ve got existing Framework, manage your production packages that are working fine. In production you’ve got a 500 reports, again assembler fairly complicated packages and they’re working, OK. Yeah, there may not be a need to port a 1 for 1 implementation.
To a data module.
It may not be worth the cost benefit and IBM has no plans to remove product support for Framework Manager.
It’s not going to be deprecated. Much like transformer which has been around since the mid nineties. Transformers as product line, it can be used with Cognos Analytics 11. The same can be saId for a Framework Manager. Now, as you look across your existing models and framework packages, there might be some opportunity to take maybe portions of it.
And there might be some packages that you might want to really port over, so that you can take advantage of some of the new functionality with data models. But I would say that you don’t want to take a complete blanket approach and say, I have to move over. Because IBM is not going to support this at this particular product to any longer. So, that’s, kind of like a general statement. I wanted to get that question off the top.
So, if you are going to be starting a new modeling task, you might want to give a serious look at data modules.
If some of your modeling tasks don’t need some of the features that are still kind of considered gaps, you might want to give it a serious look. And, in fact, some of the product managers at IBM even stated that you look at data warehouses, maybe start with data models for new projects until you get to the point where you realize that you go through your requirements. Maybe it’s not the right tool yet for the job.
So, let’s talk quickly about Cognos Framework Manager.
I’m sure that a lot of you out there who are career manager guru’s developers, have been using it for quite a long time, but, in case of some people out there who wanted to see what’s available in terms of the cross, if they are Cognos landscape, let’s talk about these quick bullets.
But, a framework manager is been the primary modeling tool since Cognos Reports, for those of you who remember that initial release of Cognos on the web than Cognos 8 in Cognos 10.
And it’s still widely used in Cognos 11 and 11.1.
In fact, there are a lot of customers out there who had migrated from Cognos 10 to Cognos 11, who really still only use the framework manager package model paradigm, where the modeling is really IT controlled and centric, and they have not even scratched the surface of using some of the newer modeling tools, like data modules within Cognos 11.
And historically, Framework Manager has been that IT centric tool used by developers and data modules.
Data mining, data modelers.
And it was never really meant for end users, you know, but I’m sure there’s some organizations that I’ve tried.
They’ve actually let end users get framework manager install on a PC and try to do some packages, But, I think, in general, some users, power users, have kind of throw their hands up in the air.
So, this is too complicated for what I want to do because I don’t need all that complex functionality of that Cognos Framework Manager has had because its development. Genesis has been more for a very centrally maintained metadata structure for developers and data model modelers.
It’s kind of like from an era from the single source of truth data warehouse, we all remember, this was the, the, really the end goal, I think, of maybe 10, 15 years ago.
We need a data warehouse is going to have everything in it that’s going to have a pure, pristine metadata layer.
And everybody will go to that for all their analytics. And as we know, now, we do need data warehouses. That’s definitely there. We do need some controlled governance on metadata and data warehouses.
But I think it’s become kind of one of your new sources, one of your sources, to newer BI analytic processes to your data scientists, and your power users, who actually maybe extract data from that data warehouse through a framework package, and then integrate it in other tools that allow them to do that data modeling, much like Tableau and Power BI.
You know, they’ve come from the perspective.
Like, users want to model, users want to integrate different sources of data. They don’t want to be hamstrung by data warehouse. Where it takes months, weeks, whatever to add more tables. When I go and get the tables myself, so, it’s a kind of a changing landscape, and it’s good to be aware of that. And also the fact that IBM has stated there will be no future enhancements for this product.
It might be a good time to start looking at when how can I use a newer modeling tools like data modules, in Cognos Analytics 11?
Now, Cognos data models itself on the converse is a web based tool.
Unlike Framework Manager, which requires installation on the PC desktop, it’s end user focused.
It allows the end user to do their own data blending.
A term that you always hear in Tableau, we can data blend. Well, as what data models can do, it’s also a good modeling tool from even basic modeling, much like, framework manager against a relational database, and there’s some transformation things you can do within the tool.
And this tool debuted initially with Cognos 11.
And it’s really IBM’s response to the whole data democratization trend in the industry.
But we need to be able to have end users, power users, data, scientists, access, to more data, and have a more agile approach to blending and data modeling.
So, it’s IBM’s response, I believe to tools that really kind of got the mindshare over the last 5, 10 years like Tableau and Microsoft Power BI.
And, for those of you who might be struggling out there with people in your organization using Tableau on Power BI and maybe just using Cognos Analytics as an extraction tool, Allowing users to use the newer tools within Cognos will help you maintain that foot footprint if that’s something you’re interested in doing. And trying to maybe standardize a little bit more on Cognos analytics and give it a lot more use across the end user community.
11.1 significantly closed some on the lot of the technical gaps between FM and data modules. So you modelers out there who have looked at data modules and 11 point no, throw your hands up in the air and say, well, I can’t do this. Can’t do that. 11.1 was released. That really focused on that. We’re going to talk about some of those things here.
And, of course, all future developmental resources within IBM will be focused on data module enhancements.
So data modules will continually get better, and a framework manager will be, will be staying exactly where it is in terms of its current development life cycle.
This slide is just really trying to show you that data modules is an integration point of many different sources. Now.
On the right are the data modules.
We basically have databases that data modules can connect to directly, and bring in tables through data server connections.
We can actually use data modulates to go against your FM packages if you need to. So if you have a lot of FM packages out there already, you still want to leverage data models for certain use case that you could connect directly to the F and packages that are out there, which will eventually go back to your databases.
And, one of the big things, of course, is uploaded files.
Your users can upload your Excel files, CSV files, into data modules for datasets, data modules for data blending, And then the new dataset feature up there on the upper right allows people to extract data from Cognos database from databases that are exposed to Cognos create subsets of data that are now stored on Cognos or faster, faster performance and integrate that in the data modules.
And what’s really interesting so data models can feed other data module. So the permutations is quite large of what you can do and it just really a matter of architecting it properly. But it allows you, that functionality is it allows you a lot of different ways to model and integrate and blend and cleanse.
So the key thing a lot of FM folks are waiting for, potentially, like, what are the what, and the current state level zero point one, release six?
What our data module is still missing that might stop you right now, anyhow, to use data models?
Well, first of all, it does not do DMR demand dimensionally model, relational models, at this point.
Now, for those of you who have never used it, this may not be a point that you care about.
Personally, I think GMR models are not necessary anymore, especially with the way a lot of the features are available in Cognos, like, navigation past, you could kind of simulate that hierarchical model.
And DMR models require a different skill set and reporting, so you’ve got a certain set of report developers, who can write reports against relational models. But DMR is another different set of skill sets you have to have.
But, again, if you, if you are, if you have DMR requirements, you need to use it. Data Module, student that still did not do that.
Object based Security, and I hope they could really fix this one up soon. That doesn’t exist still in data modules.
You want to hide a particular table, are a particular field from within the model of data models, on only exposed at a certain groups, we pay can see this table, group B can’t see that table, that doesn’t exist. So, you have to do workarounds around that. And I’m thinking this is going to be on the upcoming release. I would love to see that there. We could simulate packages better with that one object based security is in place.
I don’t know how many people out there have used team based modeling.
I personally have only used team based modeling once where multiple modelers can model against the same FM model, using model of branches and bridges.
That’s not in there. I don’t know if that’s going to be in there, But, again, I don’t know how many teams out there have actually use it, but, if you absolutely have to have that, that’s not in there.
Another Big one is parameter maps. This allows for the, is, in one particular, use case for multilingual packages, depending upon your locale of Cognos. So, it could tell you where it could it could based on where the locale of Cognos is set up? If I’m using it for English, if I’m using it for France, He could dynamically change the views of the data through parameter maps. That’s not in there. Another case where I’ve seen parameter maps users for fairly complex row level data security implementations, you might have security tables. You might have to have some fancy macro substitutions in your query subjects. And then, leverage and Parent parameter Maps For that. That’s currently not there. I think that’s on the development timeline, but that’s not in there, either.
From a style perspective, FM style namespaces packages the whole way we model on FM is different and Data Module. This can become somewhat simulated.
But it’s a different kind of thinking.
If you’re trying to do a 1 for 1, Manage the data modules is not really there, but I think some of the functionality is there, sit through some of the newer features that are there in data models, and we’re going to go ahead and discuss that.
But what do data modules do, what data models do, that framework manager can’t do. Some of, these are pretty big.
Because as the development of data module’s continues, the gap between data models versus Fabric Manager gets bigger.
The number 1 big that I can see is the easy integration, uploaded files Like Excel CSV can’t do that in a famous manager.
And a lot of times, your data warehouse or your model out there, and the databases that are centrally maintained and govern don’t have other things that users need to be able to do analytics. They might have an Excel spreadsheet with a lot of different kind of roll ups from accounting codes to job categories are things like that that they maintain. And they want to be able to blend enjoined that in their queries. So what happens is they get the extract from Cognos bring it down to Excel. They’ve got there.
Roll up file on a different tab and they do their V lookups there. Well, now, with this weekend, have end users, bring those files up, and then do the data blending within data modules there.
Simple data cleansing is also built in, for, for debt, for data modules. We can create data groups in that area.
We can quickly make Uppercase, Lowercase, and Conversions, things like that are kind of built-in to data modules.
Easy hierarchy creation.
So, navigation paths, where I could go, create my Hierarchy from Product line, the product type to product, within the model itself. So, the natural drill down is in there.
When you write reports, are due dashboards, that’s already built into data modules, Automatic creation of relative time, filters are now in there.
I have a separate webinar that kind of shows you how to do that. You could look it up on the Senturus’s site.
This has always been kind of a head scratcher within framework manager to kind of do this with complex filters, neighbors, and database changes on a time dimension. But, this automatically is done for you with data modules. Once you put that in, A plugin is fairly simple.
With the use of datasets, you can bring datasets in and, as you use datasets extracts, did extracts from Cognos itself as you as users use it? They go into in memory. So, these highly formatted parquet files will be placed in memory on the Cognos server and they kind of a job in a round fashion. But, automatically, that’s built in once you integrate datasets into your, into your data module data grouping. We mentioned, and web based interface. The web based interface, I think, to me, is one of a good thing, and a bad thing, especially for those of you who have used Framework Manager for quite a while. Because the PC is, has a lot more finer control Things, like modeling tasks. But what, the web based interface. So, you don’t have to download a program. So I think the web based interface has come a long way since 11.2 release.
But I think that particular piece of the modeling tool is going to keep getting better and better, OK.
So, what this webinar is going to address in terms of a demo, are things that typical FM modeling tasks that have been maybe more of an issue within data modules, or they weren’t there 11.
And you might want to see how we can maybe work around that, or have the new features kind of remedy that.
Um, like I mentioned, the web UI is not as precise and snappy as FM, but it’s getting better. The three demos we’re going to kind of do is the fact that FM does not have the kind of, data models, do not have the cost of the layers, and namespaces and mala query subjects.
Alias shortcuts for time dimension type role playing things and determinants, OK? So, we’re going to go and talk about that real quick.
So, namespace is a model a query subjects. So, the typical FM paradigm of layers and namespaces are missing and data modules.
No, I want a physical layer. I want to be able to hide that. How can I publish a centrally controlled data module on the right you see?
And, after working with data models for quite a Bit and I keep looking back or fabric manager, I look at it as it looks pretty complex. Actually, if I was an end user, so, those namespaces are aren’t there, but for those of us who do Favorite Manager every day, this is something that maybe stops me from wanting to do this. So, let’s go over to Cognos right now.
OK, so, this is actually, we’re running 11.1 Release six and, for those of you who haven’t seen it, there’s some cleanup that they’ve done, Some of the icons Look a little Bit different.
I think it looks a lot better if, tighten it up and we’re going to go here and create a new data module right away.
OK, I’m going to go into my Data Source Connections data servers. And I’m going to go ahead and pick the Great Outdoors Sales database that’s available.
I’m going to select my tables and I’m going to pick the order details, the order header, the product table, the product line table, and expand on that a little Bit here: the product name, lookup table, and the product type.
So, I’ve got details, header, product, line, name, and tight, and say, OK, brings us into the data modules in, if you haven’t seen data modules in awhile has been a lot of changes.
The relationship view is here and right away, data modules will inherit the joins that are in the database if the joins are defined by foreign and primary key constraints, and that’s what it has done. And you can delete them if you want.
I’m going to bring the product name. Organize this a little Bit better.
Product Name Lookup over here, Product line, product type, OK.
Then over here is Older Details and order header. So this is a typical scenario, right? Is more of a maybe an OLTP type schema basically brought in my physical tables? My database for you if you’d like.
I’ve got it over here, and generally what I want to do is try to make this schema more presentable to the user, and make it maybe more dimensional, right.
Let’s go through some of those activities then.
So, for example, I want to maybe, really just have a sales fact table.
Now, with the new release of data modules, and 11.1, we were, we were able to create custom tables.
The newest release, and I think it was five, they have custom tables over here, so, you can kind of see what they are.
But, I’m going to create a custom table across a two tables, details a header, with new table, and, I’m going to create a new view, when I create a view of the tables, it’s going to inherit the joints that are physically are recreated across those two tables.
If they’re not joined, I can create a join view here, as well.
There’s a lot of different areas where I can actually create different views with tables, but we’re going to go ahead and create a view of those tables.
And I’m going to create those beamed by just clicking Finish, and right away here is my new view of the tables.
This is kind of U 2 Release 5 custom tables. If I click on this, it shows me what those tables are made up of, OK.
And also, new to Release five is the ability.
And this was a big asked by FM modelers the query information underneath that.
So this didn’t exist for awhile and a lot of FM modelers, throw their hands up, and say, well, if I can’t see, what query is made up of that view, I can’t have fine grained control over that, OK?
So over here, it shows you the fact that it did inherit the join across those two tables. It shows you the SQL. So that’s kind of you, and that’s kind of really, what these features have been added in, to kind of addressed the gaps that FM has, versus Data modular data models has, versus FM.
OK, I’m going to go ahead and rename this table to Sales Fact.
And you can see by the icon, it’s a little different.
So this is basically a view. And what, basically, this is as a model, a query subject in a sense, create a model Query, subject across order, header and details.
And I have it over here. Now. If I could really draw a line across this here, I have my physical namespace layer. And here is my presentation layer in the sets.
Now I want to flatten these tables out here also, because I want to make a product dimension. So people don’t have to go to multiple tables. I could do a shift, click over this.
Highlights it over here, it highlights it here, right click. And I’m going to create a table. Again, I’m going to create via tables, and say next.
And I’m going to select the columns I want, because I don’t want all of these to these columns because these tables have a lot of support for multiple languages. And I don’t need all that stuff.
I guess. I just need, I’m going to go clear all and de select the columns.
I was like product number, base, introduction date, discontinued date, OK From product line.
I’m going to pick the product line curve the English name version of that from product name, look up.
I’m going to want the product name and description. I’m also going to want the product language. And we’ll talk about why I want that in a second.
And on product type, I’m going to want the product code, Type curd and the product type English name.
Do I got everything in here, Product line, everything, good things I’ve finished?
And there it shows you exactly what this view is made up of this Model query subject can assess. I’m going to rename this guy to my Product dimension.
Now one thing I want to show is the product name, look up lookup table. Then we have a grid panel where you can see that. The data, and that’s one of the cool things about it as I started working with as being of being an FM model for quite a while. The ability to just kind of look at data really quickly, is very nice. OK, you can see that we have multiple product languages for each product, and if I left it as that, I’d get double counting when I joined this to my sales fact table. So it wasn’t a pair when I edit that. When I created this View as well, where do I put that filter?
And it wasn’t really in the Create Table dialog box where that is actually over here now, Properties. So you see filters over here.
I could put a filter on that view, and it allows me to say, well, what kind of filter do you want to put as a column? Do you want me doing a column, which I do? But it also has this expression editor. Now, I’m going to click add a filter for expression editor. It basically comes up with the expression editor for reports.
I can even use macro substitutions in here. If I do want to do something more complex things for this view, OK? So that’s, that’s in there now.
But what I want to do is, I want to add a filter based on the product language.
I want to do add a filter is going to distinct on that, and I only want to filter on English.
And I’m going to say, OK. So right now, the icon now is showing you that we’re filtering this product Dimension view.
And this is the filter we’ve added. And if we again, look at the query information, we could see that as put the where clause right there, OK.
Going back to my relationships, the next thing I’m going to do is create a join relationship, a typical star schema type, Join relationships here, from product to Sales, fact, product number two, product number, match, my selected columns, and say, OK.
So I’ve done kind of the core thing of creating the start of a star schema. And within here, within here, I could much like a Model Query subject, presentation layer.
I could actually move my columns around within the product dimension here. I can rename my columns.
Just get rid of that, Ian, for example.
And I could also know Crepe folders here for measures.
Make it more, presentable.
To the end user, much like you do in the presentation layer and FM.
Like that, OK, I’m going to go ahead and save this guy, save as in my live webinar folder. I have a backup there, in case I need to, but I’m going to call this as DM two, for example.
Save that off.
And the other thing I’m going to do to make it more presentable to users, I have all these tables here which now these reviews off and Manager, I would create a package and hide the tables that I don’t want people to see.
What I could do here is create a New Folder, and we’ll call this the physical layer.
And I’m going to just pop all these physical table’s back into here.
And I’m going to go ahead and say Hi.
OK, so now I’m going to go, Dave, that one more time. I’m going to go back out here. Now you could, if you wanted to, click, try it, it’ll bring you right to the reporting tool. I’m going to go a little, some different I’m going to do, but a typical user would do, I’m going to go to my DM two. I’m going to right click on this and say create Report.
So now my data module looks kind of like a package, right?
And I’ve hidden the physical illness of that, and users can now write reports against this, any easier fashion, because my measures are over here. My product dimension is all one flat dimension.
And I want to, maybe, product line by, product type, insert that. I’m going to insert quantity.
And then run that query.
So there we go, OK.
Let me go back to my data module.
So we’ve, we’ve done that. So let’s go back to that, what we were talking about. So namespace is the mala query. Subs, OK? It’s not a 1 for 1, right? But it’s kind of a leaf. We’ve kind of fill some of the functional requirements to maybe do something like that.
The other option, as I say, how can I publish a simply controlled data module?
What some people might want to do is crater on data modules but they don’t want to know how to bring all the tables and do all the join from a physical perspective. But we can do is maybe just publish out a data model that only has a physical layer in it. It’s it has everything there. We might put a read only stamp on that and then people can work from there. So, if I go over here, let me go to.
This folder, I have a physical layer over here.
I could actually create another data module off of that.
These are links.
The end user will have no way to change that.
But then I could leverage the ones, the links here, that IT has created. And I could then create my table over here that I can create my sales fact this way.
This becomes my sales fact folder or query subject, model query subject, OK, another alternative way to do it, uh, in publishing centrally control type IT maintain data modules so that some of your more power users, your data scientists, whatever. Can use those. And create their own data modules and build their own presentation layer on top of that, OK.
So as you saw, one thing that I kind of glossed over, I kept changing this and saving it. And what’s the reporting to the right away? I didn’t have to publish a package, right. I saved it and published the package, I save, than I went to the reporting tool. I wrote my own reports real quickly.
So, the Prototyping the Agile ness of the tool is much faster than, obviously, framework manager, where I have to model everything on the PC. I’ve got to set up my package.
I got to publish it, you know, and then I go back into Cognos, then I look at the package and, you know, that whole thing is all within the web tool, OK?
That’s another differentiating factor.
OK, let’s go back to our slide deck real quick. The next one here is the concept of alias shortcuts. This is a very important piece of modeling the modelers love to use. If you look at the framework, manage your package, a model of there. We have the sales fact. They’ve got two dates over there. They have the order date, and they have the ship date, and I need to be able to analyze across two different time dimensions. I got one physical time dimension table.
And in Fabric Manager, I was able to right click and create alias shortcut off that physical time dimension, and use that and leverage that, OK?
There is no right click Create Alias shortcut in data modules. But let’s go back to data models again.
So I’m going to bring into my time dimension.
Over here, Add more tables, I’ll look for it.
Again see, the search is pretty nice, is pretty quick time dimension, and say, OK, there’s my time dimension. The one thing I also mentioned, as aside from a maintenance perspective, released six got really good to happy. From a usability to maintain these modules, because what we didn’t really have before as a way to reload the metadata to see what tables we have in our data model versus the physical tables.
Here in our sources, we see all the physical tables that are available. I’m going to go on my time dimension over here.
And I’m going to maybe delete some of these fields. I don’t need these fields.
From the physical table that I brought in.
There’s a new feature in here, which shows unused items.
And if I expand my time dimension over here, scroll down.
These aren’t there, so if I did, not, if I deleted them here, I can actually, now, individually, from a field perspective, just drag it in before there was a real pain to kind of get that done. And also before, somebody, physically, if the DBA changed a table, and added more fields, it was hard to refresh. That had to go back to the data source connection and refresh it. Well, now you have a right click here and you can see reload metadata.
OK, so another FM modeling issue was, Alice too hard to maintain the field’s command? It’s really hard to bring in a new fields on that, OK, so that was another thing that was kind of fixed. And basically major your life easier to use.
All right, so back to the original thing, the alias shortcut. I have a time dimension over here.
Aye, want to be able to create alias shortcut? Well, it’s not there but we can leverage this whole concept of custom tables. We’ve already created a couple of these. We can go to the time dimension.
Say, you table.
We’ll create a view of the table.
And I don’t want all these fields, for the purposes of this demo, I’m going to clear all, I want only the day date. I want the month key. I want the current month.
I want the core to key our current quarter, I want the current year, and maybe I want the month, English name.
I’m going to bring that in.
Now, I’ve got a view. Now, my views are kind of building up, I have another view over there, OK.
Um, and the one thing about this, and brought the current month number, and current core number and as Measures, by default, because it’s in America.
Variant Manager had a really nice feature where if I did that and drag down the attribute column and it’ll change that, thought, we have something similar here as well, I can multi select that column over here, go over to my properties, pain, which is over here.
And change the usage over here to say attribute, which is what I want to do. OK, and I want to make this, the ship dimension.
So I’m going to, I’m going to rename this, the ship date dimension.
Now I also want to be able to have an auto date dimension here as well. I could do the same thing, I could create a custom view of time dimension. But it’s a really nice copy and paste type feature over here. I’m going to go ahead and paste.
And I’m going to rename that guy to be order date dimension.
OK, then I can rename my fields, making them more pertinent to be whether it’s ship or to date like I would want maybe.
This one here to be, you know, rename this to be the ship date.
Etcetera, I want that to be the ship year.
Same thing with Order Date Dimension. I want this to be the order date.
And I want current year, to be order you.
These are my roleplaying dimensions.
Ship Date, OK, get this filtered out of the way.
And, I want to now just simply join ship Date to my Sales Fact.
I’m going to clear relationship from here to my sales fact: Ship, date to ship date.
Or a date relationship, too.
Sales fact, or due date, or a date match?
We’re good, OK?
So, you can see my star schema is getting linsey built out over here.
I’m going to save this.
Also, what’s a nice feature? I’m going to go back to my reporting tool to test this out.
All I have to do is refresh loads backs back up.
And I’ll bring it back in.
Well, it should have.
So I did that, say, uh.
Say that again.
Let me go back out here and OK, I did bring it in. I just wasn’t waiting long enough. There we go. So there’s my ship date dimension.
And I’m going to go over here, delete that.
I’m going to do a list.
We’re getting an error as heroes.
I’m going to close this report down.
Going to go back over here to my live webinar, Enter, or write that report again, Create report.
OK, now let’s dial list.
OK, so my ship date, I’m going to bring in a ship beer.
I’m going to bring in order here.
And I’m going to bring in for my sales fact.
Order quantity, OK, and I’m going to go ahead and run this.
I actually should have put order here. Let me do that one more time. I’m going to actually, put order for us to make it easier to look at.
Order year! Over year.
All right, let’s rerun that again.
So I’m basically now running a query across two roleplaying time dimensions, so as you can see, things that were ordered in 2010.
Had things that were shipped in 2011, the next year. And things that are ordered in 2010 had things that were shipped, and also in 2010.
OK, same thing across here, so that shows you the fact that I’ve been able to slice across two time dimensions with that data module.
So going back to the data module.
Let’s clean it up again, and I’m going to put the time dimension back into my physical layer.
I’m going to go back over here.
Now, guys, ship date order, product sales, save.
Let’s test the Cognos gods and go back to that, you report and refresh that and see how fast now? It was faster? That was good, OK?
By the way, our instance of this is on Azure. We have Cognos on Azure installed so I am going to go into the cloud to run this. It’s pretty snappy, as you can see just a little side note, so I made my change the data models real-time. I saved it. I went back to my report refresh data and now I’m building out my model here. Now, I’ve got to ship date, order date, product dimension, sales fact that I’ve done fairly quickly in this modeling exercise.
Trying to address some of the key issues as you move from framework manager, how I do, for example, uh, roleplaying dimensions and namespaces, et cetera.
The last topic we did alias shortcuts, custom tailored views, and it’s always been kind of a head scratcher and but for those professional modelers are those Mahler’s power users and know how to do this and framework manager. It’s a really important feature to have. This feature did not exist in Cognos Data modules: Data Module 11.
It now, it started to exist and 11.1 and has done better for the releases, determinants, manager, versus columned dependencies.
We need determinants when we need to use one time dimension, as say, the day, grain, and we want to use that same time dimension and cross multigrain facts.
So, for example, I’ve the sales fact that I’ve created is at the date granularity, but I have sales targets at a month level.
I want to be able to do cross joins across those things across the two for cross fax multi queries with the same time dimension.
In Fabric Manager, we use determinants, and here’s the screenshot. And every time I don’t do framework manager modeling for awhile and I have to look at the determinant screen, I always have to scratch my head and say, how does this work? Again?
It was never really that intuitive.
And if you can imagine, end user looking at this screen, they would, again, probably just throw their hands up in there and say, that this is too hard, right? And that’s what we’re trying to get away from. Would not want that? We want this power to be enabled in the end user’s hands.
So Data module’s has simplified this concept with the concept of column dependencies.
So let’s go back over here, and I’m going to bring in another table.
Add more tables, select tables.
And I’m going to bring in my, look for it, look for it. Sales target, here we go.
It inherited a relationship. I got to get rid of that for the purposes of this demo.
Here’s my sales target.
If we look at this data from the Grid view, we have sales targets over here, at the grain of Year, month. And even at a higher level of product.
We could put determinants on product as well.
But what I want to do, and, again, for the purposes of this demo, I want to be able to use one time dimension, we’ll say, ship date dimension across the two, Right, if I were to create a relationship to sales target right now, and the grain of sales target is year and period.
And I went from ship year to sales year and current month to sales period, Mats selected column, and say, OK, right, and say I’m done.
The effect is this, I’m going to save this go back to my report, Refresh.
Sales target is now there.
We’re going to delete this real quickly here.
We’re going to create a two column report, over here.
I’m going to create a list, and I’m going to query my sales target table directly.
I’m going to bring in these three guys.
Over here, I did a join to my ship date dimension.
And I’m going to leverage the date field from ship date.
I’m going to put in that **** beer.
Click off of this. I’m going to put ***** Beer.
Current month number.
And I’m going to put in that sales target, and start over there. Let’s run that query.
Alright, so this left one is from the Ross, the raw sales fact.
Table by itself. This is the right number.
Sales target for January’s, $57 million.
But, over here, the sales target was inflated 31 times because the date dimension is at a granularity of the day. So for January, multiply 2 times 31, February maybe 28, et cetera. So how do I fix that is the question.
Let’s go back to Data Modules.
I need to be able to create a determinant on the date dimension.
And with data models, we now create column dependencies. If you right click on here, there’s an option called Specify, Column, Dependencies, and this is really easy.
So, for example, I’m going to start with ship year. That’s a top level.
And I’m going to bring in the key for quarter level is the quarter key.
And all you have to do is drag the relationship here. So ship goes to Quarter.
Quarter goes to month, and drag it over here.
Month goes to Day.
And one thing about data modules, it does a constant data validation, or validation of your model. And right away, it tells me there’s something still wrong.
If you click on this validation issue, and you show the details, what it shows you is that there are certain columns in your column dependencies that you have not accounted for, current month, current quarter, month, English name. If you actually went and query included, those who wouldn’t know what the dependencies are.
So, all we have to do is clean that up. So, current month.
Number is really an attribute of month key. We’ll put that over here.
It really validates and the list is getting smaller.
Current quarter is an attribute over here.
OK, and we’re down to month end, which is the month, English name or the short abbreviation, and that’s going to go over here, and the validation will go away.
I’ve established my column dependency, and there’s also a vertical view you can look at.
However you feel, it’s most useful, OK, so now, this relate this, this. This date dimension is now properly set up.
Let’s save it.
I’m going to go back into my report.
Refresh the data.
Let’s rerun the report.
I was hoping you would come up with the right answer, and I did this being a live demo.
So you see now, even though I’ve used a time dimension that’s at the day grAIn, across a fact table, that’s at the month green, I get the right answer.
OK, and now I can make proper queries with sales Face to get me a multi grain, a multi fact query. I’m just going to use Quantity for this example.
I’m going to put it over here, and then, run that. It’s not going to be really a good comparison of Sales target, MLIS. I just wanted to kind of show the query information on that.
OK, so that works, what does a query look like for that?
Lot of query, But what I want to show you is, it has the Coalesce statement right here.
That’s what it’s doing.
Two queries against two fact tables, using a date dimension, that a different grain, then one of the fact tables, OK?
Go back to my data module.
OK, so in the span of, really, actually, if I were to do this without talking, and I knew what I was kind of doing. This is really a 15 minute deal. Right?
And I’ve kind of addressed some of those issues with Fabric Manager, Right? So back to my slides, over here.
To me, that was a lot easier than trying to figure this out, and then publishing the package.
As you can see here, there’s a determinist uniquely identified group by, what are the attributes for it? Very complicated, from an end user perspective.
So those were the three things I wanted to demo today. The takeaway from this webinar, as, I hope, that you can see that the gap between Data Module modeling from a professional modeling perspective. From those folks, who are you constantly used to using framework manager and creating a package?
That Gap is definitely narrowing release six with that refresh metadata, but there’s a simple button, but I remember I was working with customers and saying, until I can do that, I’m not going to touch it because I don’t want to be able to delete tables and bring things and all the mast and redo everything. There’s no way I’m going to put that out into production. I can’t, I can’t maintain it, OK. They fix that. That’s done, and that was a huge thing. Again, this webinar was kind of tailored to the FM folks and to show you what you can and cannot do.
You’re going to have to make your decisions on whether you want to start that new modeling project with data modules. I think you should definitely give it a hard look at first.
Before you think about maybe starting a brand-new project with framework manager packages, you go through the list of what it can’t do.
If you have slightly habitat parameter maps for link languages, complicated security models, you need object level security, for whatever reason, and I think that’s going to get fixed soon. Then, you go, you’re going and you have a deadline. You’re going to have to go back to your fabric Manager packages, you know, but it allows a lot more functionality and we didn’t really talk about the functionality that end users can use us for. This webinar was more geared for framework manager versus data models. But, there are probably a whole set of customers out there who are bringing coming in, maybe Cognos Analytics, 11 clean without any of the legacy history of it.
And using data models. Because that’s the modeling to know. And even consultants out there, you’re going to have to maybe go into customer bases where they’re going to maybe say, I don’t want to use Frank Manager and for whatever reason, to use a data modeling tool. But you have this correct, or you’re the kind of still leveraging your knowledge on…
Manager, and you want to be able to take those best practices that you’ve learned framework manager and possibly. Go and use it in the new modeling tool the modeling tool.
That’s going to be enhanced and all development resources are going to so that is I was able to do it.
At 11 55 I was a little worried about was a fairly complex demo with Pedro that up so I’ll get back to Mike.
We’re moving going to move pretty quickly here. We might run a little past the top of the hour mm stick around, folks. So there’s a there’s a million questions, Pedro. So go ahead and take a look at those projects working from the top-down. If you don’t mind random past the top of the hour, people will get into those questions in just a few minutes before. We do a quick poll in terms of asking you going forward. What do you plan to do in terms of Cognos metadata modeling? And you can select all that apply. So are you going to use framework manager data, modules something like Microsoft Analysis Services, or something else?
So go ahead and get your votes in there.
Give me all about Nan to do this, since we’re on pretty tight time here.
Great. Kind of a two thirds, you getting your votes in here. That’s great. Alright, I’m going to close it and show it. And so it’s still two thirds, and then a whole ton, almost 83, 83% here using data modules. And than a smaller percentage of Microsoft, and other undecided.
All right. Thank you very much.
So, before we get to the, if you could advance the slides, a couple of them, Pedro, to the gut cognitive challenges. So in your organization, one more if you have if you’re suffering from time consuming data prep, where you can’t drill down to from summary to detail, you have performance issues, or for user adoption, we can help you with an architectural assessment that we do, that involves best practices, modeling.
We do a health check on your metadata, performance tuning, tips on enabling self-service culture, and selecting the best tool for the job, for example, using framework manager and or data module’s as appropriate.
So a little Bit about send tourists here. We are the authority in business intelligence. We concentrate our expertise.
We go to the next slide, Pedro, on business intelligence with a depth of knowledge across the entire BI stack and next slide, please. Our clients know us for providing clarity from the chaos of complex business requirements, disparate data sources, constantly moving targets, and ever changing regulatory environments. We’ve made a name for ourselves because of our strength and bridging the gap between IT and the business, by delivering solutions to give you access to reliable analysis ready data across your organization, enabling you to quickly and easily get answers at the point of impact, in the form of the decisions made, and the actions taken.
Next slide, our consultants are leading experts in the field of analytics with years of pragmatic world expertise. And experience advancing the state-of-the-art, everything from front end dashboard, reporting, and visualization creation, through Data prep operation, all the way through infrastructure, and, of course, training and mentoring.
We’re so confident in our team and our methodology that we back our projects for the 100% money back guarantee that’s unique in the industry.
We’ve been doing this for awhile, but at it for almost two decades, Focus exclusively on business intelligence. We work across the spectrum, from Fortune 500 to mid-market companies, solving business problems, across every industry and functional area, including the Office Finance, sales, and Marketing, Manufacturing, Operations, HR, and IT. Teams both large enough to meet all of your business analytics needs, and yet small enough to provide personalized attention.
Again, we invite you to visit the Senturus’s website, the Senturus Resources tab, where you can find all our webinars and the recordings on the deck and everything behind those until all kinds of great information. Our blogs and tips and tricks and everything on that site bookmarked upcoming events. We’ve got another great webinar coming up on Tuesday, July 28th, same time, same Channel, Cognos analytics dashboards, or reports. So, come in.
Visit us for that. And, of course, we would be remiss if we didn’t talk about our fabulous training.
We offer training in the top three, three top BI platforms, IBM Cognitive Analytics, Power BI, and Tableau, right? Deal for organizations running multiple platforms, or those moving from one to another. And we provide training and many different modes, It can mix and match those to meet and suit the needs of your user community.
Senturus provides hundreds of free resources on our website, and we’ve been committed to sharing our BI expertise for over a decade.
And now we finally get to the illustrious Q and A and so as you can see, Pedro, there’s a whole ton of questions here. I know, are at the top of the hour, so I don’t we try to bang out a few of those, and anyone, you can continue to put questions in the question panel here.
We do save that question log, and we will answer any questions we’re not able to get to, via that login will posted, along with the recording and the deck on senturus.com, and that Resources section.
So go ahead, Pedro.
If we want to bang out a few of these, we can we can do that. Yeah, and stay on the moon.
See, some can expand this one box, but maybe you could address this. Mike is IBM at least considering, I think. It’s a migration tool.
I don’t, I’m not aware that they are It’s certainly something that we in our labs are actively pursuing because we know that it’s, it’s, it’s feasible. So we’re looking to do exactly that to help people if they want to migrate from Framework, Benadryl models to data modules and have some technology to help them with that.
So if you’re interested in that that’s a need in your organization, definitely give us a call OK, can data modules reads synonyms? Well, that’s an Oracle question, I think, more specifically as you connect to an Oracle database through the data server.
If it’s part of the schema that you’re trying to get to my initial gut, is it does.
I’d have to actually verify that and look at it from an Oracle database and see that it can read a synonym, but if the user that you are logged into, can see it, I’m pretty sure the data module can see it.
Procedures, stored procedures? That’s a good one. There is a widower, right?
I haven’t tried it was stored procedures yet. But you can write manual SQL and if you can make manual SQL, you can do a create table and I didn’t show this one.
Create a manual table with, create a table, with Manual SQL. You could potentially, maybe, call a stored procedure, but I’d have to try that out, that’s a good question, sure, Sure, and then, someone mentioned about DMR and framework manager can. And I know you can create hierarchies and data module. So that’s kind of similar to that functionality, but not exactly equivalent, right? Not exactly equivalent. I mean, the EMR is very specific corner section of bench they can do, which makes it look exactly like an old … cube. That kind of functionality is not built into data models. Whether it goes in there or not, I’m not sure. We haven’t got to read from that an IBM yet.
OK, we had a question about Automating the Refresh, the data modules, I’m not sure exactly what that means, necessarily, because they’re not really.
Yeah, I think maybe it might. Unless you have less, you’re referring to data, sets, extract data that can be put into data modules, those things can be automated to be refreshed, because you can set the set those on a schedule.
It’s very easy to do from the user interface.
You create your dataset, and you set a property about a schedule, you enable that. And then when, when people hit the data modular, which contains a dataset, it’ll show the most refresh data. So, that’s the only thing about refresh that related to data models. Yeah. Yeah, exactly. Someone asked about the Senturus’s analytics connector and if it supports data module. So for those of you an initiated, we have a product called the Analytics Connector that allows you to use Tableau or Power BI to access your cognitive metadata. So, Framework Manager Models, reports, and Data module. So, they asked if we supported data modules. And the answer is, yes, we do.
Filters on the FM, Database Query subject, based on session parameters. Can you replicate that in data models? And I’m guessing that you can’t do that because of that. We don’t support parameter maps.
We don’t support parameter masks what I haven’t tried out. It does create a filter with an expression editor. And I’ve seen that it can reference macro substitutions.
Um, So that’s something to look into and try.
We’ve done that I’ve done some very complicated thing was parameter maps Macro substitutions and FM, like getting to security table and just in the where clause with session parameters, based on your ID, the user IDs, and the security table, etcetera. I would think that’s going to get better. And I know parameter maps are number one thing. They want to add data modules, but you might want to take a look at the area of filters, where you do a filter by an expression editor and see if you could use the macro substitutions to get. Maybe this has parameters there.
And try, try to play with it there and see that that, and that, that satisfies your requirements, OK, we’ll do one more here, and then we’ll probably to let people go here, shared data modules, Can they do the impact analysis like FM does, in terms of determining and showing you impacted reports by changes that you make?
No. Not that I’m aware of. Right. So, yeah, you’re right.
If data modules are being used by a bunch of reports, that particular functionality is not there OK. Great. Well, again, we have a ton of questions here. We will save the question. Login will answer that posted to the website, if you want to go to the last site. The last slide, Pedro. I want to thank everybody for joining us today.
And thank our speaker, Pedro Ining, for a very informative and information filled session. On behalf of myself and Senturus, thank you for joining us. We look forward to seeing you on one of our upcoming events knowledge series. Or you can reach out to us for any of your business analytics, needs, your training at senturus.com.