Lesson 1 – Get Up & Running
Lesson 2 – Creating Models & Harvesting Metadata
Lesson 3 – Stitching Metadata to Tracing Data Lineage & Impact
Having OEMM installed and configured in our previous lesson, time to jump into “harvesting” some metadata. In this lesson, I’m going to teach you the needed steps to harvest virtually ANY metadata. As I’m trying to make you master OEMM step-by-step and keeping things simple, this lesson will only focus on creating new models and harvesting metadata. More advanced functionality (which are plenty) shall be discussed in later lessons. Please note that there are no prerequisites (in terms of knowledge) for this lesson.
Before we begin, let’s first describe certain terms that are used in OEMM.
Harvesting is the process of “reverse-engineering” or importing the metadata of a specific metadata provider/technology. This could be virtually any metadata such as OLTP database, ETL tool, Hadoop project, BI solution, etc…
Model (in OEMM) is the result of harvesting/importing. It could be OLTP database, ETL tool, Hadoop project, etc…
Bridges are components that are shipped with OEMM when you install it. They are provided by “Meta Integration Technology Inc.”, an Oracle OEM partner. Bridges are the interface between OEMM and metadata providers. For example, you’ll find a bridge for Oracle Data Integrator, Oracle BI, Microsoft BI, etc… Bridges may depend on the native tool connectivity component, such as ODI SDK when connecting to an ODI repository, in order to import (or reverse-engineer) the metadata. To check the supported technologies by OEMM (the bridges), please refer to the product matrix on the product page.
Now we understand some basic terms, let’s jump into action.
Warning: the complexity of this lesson is incredibly low.
Creating new Model & Harvesting Metadata
Models are saved in OEMM Repository, which can be explored from the left side pane:
To keep things nice and clean, let’s create a folder inside our OEMM repository. Right-click on the root node (Repository), then New=>Folder:
Enter a name for the folder (I used “Folder1”) and a description if you wish, then click on “Create”:
Now let’s create a new Model that connects to an Oracle OLTP Database. Right click on “Folder1”, New=>Model:
In the “Create Model” window, enter a name for the model (I used “Oracle DB Model”), a description if you wish and then select “Oracle Database (via JDBC)” in the “Import from” drop-down list:
Note that you’ll find all the supported models by OEMM in this “Import from” drop down list. Those are the “bridges” that OEMM uses to connect to various metadata providers. Now let’s click on “Import Setup” tab to configure the bridge connectivity parameters:
Note that each parameter has its own description in the “Help” pane on the right side. Simply click on any parameter to see its full description and an example of the input. For each metadata provider, the setup screen parameters will be different.
You may click on “Test” to make sure that the connectivity to your metadata provider is successful. Once you are done configuring, click on “Create”:
OEMM will ask you if you’d like to import a new version. If you wish to do so at that moment, click on “Yes” (which I did here):
If you choose “Yes” as I did in the previous step, you’ll see the progress of importing (harvesting) the Model which may take a few seconds to few minutes depending on your server resources and what you’ve selected to import. Please note that you do not have to wait until the import is finished, you may close the “Log Messages” window and OEMM will harvest the Model in the background while you do other things in OEMM. In this example, I didn’t close the “Messages Log”, and waited until it was finished. If the import was successful, you’ll get a message stating that and asking if you would like to open it. I choose “No” then clicked on “Close” as I wanted to import another Model. But you may choose “Yes” if you’d like to open it.
If at any moment you would like to see current operations and logs inside OEMM, simply right click on any object (Model, Configuration, Business Glossary, etc..), folder or on the “Repository” root node itself and choose “View Log”:
You should see something like the following:
Now our “Oracle DB Model” has been imported (harvested), let’s try harvesting Oracle Data Integrator.
Same steps like we did before, except that I’ll choose “Oracle Data Integrator (ODI)” in the “Import from” drop-down list:
Now let’s navigate to the “Import Setup” tab and enter the required information in order to connect to the ODI repository and harvest it:
Nothing really special here, we used the Oracle JDBC driver to connect to the ODI repository (which is Oracle DB in this case), and we pointed to the ODI home directory, as we need ODI’s SDK to harvest the metadata properly. This allows us to select which “Content” to import (projects) as you see in the following image:
I’ve mentioned this before, but to know exactly what you need to enter in each parameter, simply click on it and you’ll see help with an example on the right side. Once you finish entering all the required information, click on “Create”, when prompted to “Import a new version now?” click on “Yes”:
Same as importing our Oracle DB metadata previously, you’ll see the “Log Messages” window which you may wait until it finishes or close it and have OEMM harvesting it in the background while doing other things. In this example, I waited and OEMM confirmed the successful import and asked me if I want to open the imported Model. I choose “No” and then closed the “Log Messages” window:
So far, I’ve harvested Oracle Database Model and Oracle Data Integrator. Let’s now harvest Oracle Business Intelligence Enterprise Edition (OBIEE) by executing the same steps as we did before, but selecting “Oracle Business Intelligence (OBI) Enterprise Edition” in the “Import from” drop-down list:
In the “Import Setup” tab, enter the required information to connect to the OBIEE repository:
Nothing really complicated here: the OBIEE server address, credentials, etc… The “File” parameter refers to the OBIEE RPD file, which is where Oracle BI tool natively stores metadata. If you are uncertain what you should be filling any parameter value with, refer to the help which you can see if you click on any of the parameters. Now “Test” your setup, then click on “Create”. Follow the same steps we did in importing other Models previously.
The OBIEE, ODI and Oracle Database content, installation and configuration are out of scope for this tutorial. I will not go through what exactly we are importing and for what purpose, this lesson is meant to show you how (simple is) to harvest any metadata (in general).
Last model I wanted to harvest in this lesson is an Excel file, same steps as before except that I’ll be choosing “Microsoft Office Excel” as an “Import from” option:
In the “Import Setup” tab, I’ll select an Excel file (make sure to read the “Help” to understand the Excel file specification requirements):
Click on “Create” and choose “Yes” when prompted to import a new version. Once the import is finished, choose “No” when prompted to open the imported Model.
We have just finished importing/harvesting four different Models: OBIEE, MS Excel, ODI and Oracle Database. I won’t harvest any more at the moment as my intention for this lesson was to demonstrate the concept of creating new models. The same steps apply to creating models for other metadata providers, only the “Import Setup” part would be different. This is how my OEMM Repository looks like at the moment:
In this lesson, we have learned the steps needed to create a new model and import (harvest) a new version of it. While I did show you examples for ODI, OBIEE, MS Excel and Oracle DB, the purpose was to show you that the steps are the same every time except for the setup parameters for connecting into a certain technology.
As you have also noticed, OEMM can harvest from files that are stored on storage media and also via connecting to online repositories. OEMM has the capability of harvesting virtually any metadata, whilst this lesson demonstrated how to import a new model, we have also managed to prove its ability to harvest Oracle Business Intelligence Enterprise Edition, Oracle Data Integrator, Oracle Database and Microsoft Excel.
Join Oracle Enterprise Metadata Management group on LinkedIn to keep track of new announcements, articles, lessons and more. If you’ve missed any of my OEMM previous post, you may want to click here.
If you have any questions, please do write in the comments area.
Having this lesson completed, you should be ready for the next one which I should share in the next few days. Stay tuned!