Thursday, April 2, 2015

FDMEE Basics : Getting Started with this tool



Recently, I’ve been too busy with mediocre tasks like meetings -with old and powerful men- which are usually futile, project health assessments of projects which are already in good health, giving Installation advisories to people who won’t bother to read the Installation documentation even once , Application related consulting which is all over the net etc.

Bored, I thought of learning something on my own, just like ODI.
Now, I had done some FDMEE review in a project (basically, ask an expert to show me how it works) earlier but never quite worked myself. I readied a 11.1.2.4 instance and got started. I will be pasting my experience here, hoping it helps for beginners like me. 

Note: This is a very basic starter kit for disgruntled employees like me who are not given any training and have to learn things on their own (just kidding). We are not going to ‘deep dive’ in FDMEE, rather just do a small ‘snorkelling’ session which will clear the understanding of this tool and a logical flow of how things are done. I am going through the admin guide while I post this so not sure if all steps I post below are absolutely correct.

So, experts can give me their valuable inputs!

Our source is a Flat file (test crap) and target is a classic planning application (some PoC application created ages ago)




FDMEE Self Training : Level: Basic
To start with FDMEE, we first need to understand the flow of things. Similar to ODI, there is a set path of doing things in a logical sequence, geeks call it ‘Algorithm’. The first one being ‘Setup’
In this session, I am going to take Source as a flat file. Now observe where we are:
- Data Management -> Setup -> System Settings -> Profile Type: File
- Application root folder path to be specified should be on the server where FDMEE is running
- Create Location Folder is selected as True  
- Do not forget to click on ‘Save’ every time



Next Step is to create a file source system on our FDMEE.
I’ve kept the ODI context as GLOBAL (in caps) as that is the default one. You need to have some ODI understanding for this, but those who don’t know ODI can safely keep it as GLOBAL. Just remember that ODI keeps all codes in capital letters













Select source system as File. Observe the available options in Source, you will not see ‘EPM applications’ in it. Why? Read on to find out.
This step can be compared to setting the Physical Schema in Topology manager of ODI. The only difference is, here we are doing it only for Source Systems. Does that mean EPM applications cannot be a source in FDMEE ? Definitely not. In the next few steps, you will get the answer



After setting up Source System, we move on to Target Application, observe the subtle difference, Source is a SYSTEM while Target is APPLICATION. I click on the “Add” button here, and select Planning. (Note the target applications: all are EPM )



I then select the desired Planning application.
Note: This page takes some time to load, if it does not, you may want to clear browser history, cookies and offline data.


Once the application is selected, the details are automatically populated. Now, for beginners, we will only focus on some parts of this page. Observe the column headers under ‘Dimension Details’
Dimension Name : The Dimensions present in your planning application have manifested here.
Target Dimension Class : FDMEE has its own way of segregating Target application’s dimension into various classes like : Account, Generic, Entity etc. It automatically maps these classes to Target dimensions. We will leave it there as it is.
Data Table Column Name: Don’t think about it much right now. Leave it as is.
Note: You can also add more rows if you want to do some lookups or filtering. More on that in later sessions



Then comes the most important page: Import Format
In ODI terms, this is the place where you build the interface, i.e. Map source columns to target columns. Click on Add button to start.
Before that, you should see that I have kept a test file in the folder which was specified in the first step of System settings.




Click on Add button (I have down scrolled a little ), You will see that ‘Details’ area needs to be populated.
Here is the answer to the question I asked earlier. ‘Why can’t I see EPM applications in Source System selection? Can’t I use source as EPM?’
Well, you can. See the option of Source Type. It can either be an
ERP (File, Relational table etc) or
EPM (Planning, Essbase etc)
If you select ERP, whatever systems you had added in Source System step will be shown.
If you select EPM, whatever applications you have selected in Target Applications will be shown.
You need to click on Save before you proceed any further. If everything is fine, the Mappings area will get populated



Now, everything will be blank here. Click on the Build Format button. A file browser will open. You will see the file I had pasted in the server. You can upload a file onto the server from your desktop as well. Select the desired file and click on OK 


Let the Mapping Games begin !!

Select, as in, highlight each header one by one, once highlighted, click on Assign Dimension button, Once that is done, enter the source dimension name (anything logical) and select the Target Dimension it will be mapped to.
Also make a note of the field number (column number) to see if you are not going wrong anywhere.


Do the same for rest of it. Ignore columns which are not required to be mapped.

The Trap : I want to map period dimension in source to a Period dimension in target. But wait, I cant see it in the list... hold on...
Oopsie daisy ! FDMEE does not allow me to map Period, Scenario and Years dimensions at this point (Import Format window).
So how do I map it ? We will see it later.
Right now, we will just map all columns.  And yes, Map Data column to Amount (I added Data column to the file later on)




Once the mapping is done, click on Save button.
We will keep the ‘Expression’ tab for a later day.  Right now we will only focus on understanding the flow.


Location Tab: This is the place where you set the Source-Destination combo. So, whenever you select ‘Location’ going forward, it means you are selecting a predefined combination of Source and Destination. In my case it is a Flat file and Planning application called ‘AIPOC’  (Dont bother about the name)

In this window, we select the ‘Import Format’ that we had created earlier. The Target and Source get populated automatically upon selection. Leave other sections as it is.



New ‘Location’ can be seen created. Dont forget to save.

Period mapping: Remember I mentioned that FDMEE does not allow to map Scenario, Year and Period dimensions in ‘Import Format’ section ? Well, it  has a dedicated window for mapping Periods and Years.
For beginners, let’s focus only on the basic Period mapping, In global mapping tab, you need to specify:
- Period Key: The last day of the month to be mapped
- Prior Period Key : The last day of the month before the month to be mapped
- Period Name: Hard code it as desired.
- Target Period Month: Hard code, should match the destination (Planning app) dimension member
- Year Target: Hard code, should match the destination (Planning app) dimension member
Ignore the ones which are blank. The ones mentioned above are mandatory.
This exercise is to be done for all possibilities of your calendar. As an example, I have put only a few.
As of now, I am not sure if there is any other way of doing this but manually entering every combination is a time consuming activity.
Save each entry, else you will lose all the data even if one entry is wrong.
You will have to repeat the same in Application Mapping tab as well.




Next, we come to the point where we need to map the Scenario dimension members (Remember we didn’t map Scenario in ‘Import Export’). Pretty simple, just go on adding rows and mentioning Category and Target Category. The Target category should match the exact text of the Scenario dimension member.
Do the same in Application Mapping tab as well.

 

This marks the end of setting up Interfaces and basic activities for running a load job.
To summarize: These are the steps we follow
1.       System Settings: Specify file path and folder location
2.       Source System: Specify Source (file in our case)
3.       Target Application : Specify Target application (Planning in our case)
4.       Import Format : Specify mappings from Source columns to Target dimensions (except Year, Period and Scenario)
5.       Location: Create a pre-defined combination of Source-Target based on Import Format. This will be used later in workflow.
6.       Period mapping : Specify Year and Period names for Target
7.       Category Mapping : Specify Scenarios for mapping in Target.