Oracle Data Integrator: What is a Repository?

by Brent Dayley, Senior Instructor at Oracle University


While teaching Oracle University classes on Oracle Data Integrator, I often instruct my students that it takes roughly one and a half days to understand the setup that we do in ODI Studio in order to start to map data from source to target. I start out by giving them a visual tour of the four navigators: Topology, Security, Designer and Operator. After giving them the visual overview, I begin teaching about the tasks that need to be done in the Topology Navigator to get started mapping data from source to target. The Topology Navigator includes a number of different sections that are used to describe the Information System that will be used for source, staging and target tasks. Among the different sections of the Topology Navigator is the “Repositories” section:


pic1

Often I get perplexed looks regarding the Repositories and it sometimes seems to cause thoughts of “Black Magic”.


Simplifying the Mysterious:

Simply put, the Repositories are nothing more than ODI objects that map to storage areas that contain database objects, such as tables, that hold information about what we are doing in ODI. At the very least, we need to create two repositories, a Master Repository and a Work Repository, and usually a modified Work Repository for production purposes called an Execution Repository. Even though these repositories are described as two separate ODI objects, they can point to one or more actual database schemas:

pic2

Proprietary Database Required?

No, ODI does not require a proprietary database engine type. The repositories can be Oracle-based, or a number of other database types:

pic3


ODI offers flexible methods for creating the database objects that store the information about what we are doing in ODI Studio and we do make recommendations regarding which is the preferred method when setting up the repositories with an Oracle database. One of the repository creation methods is to use a separate utility, called the “Repository Creation Utility” or more commonly referred to as “RCU”, to create the repositories. The RCU is the preferred method when creating repositories in 12c, because of the auditing infrastructure, however ODI Studio also has Master and Work repository creation methods built in. When using RCU, both the Master and Work repositories will default to pointing to the same storage schema in the database. When creating the repositories using the ODI Studio methodology, one can use the same or separate database schemas to house the repository tables. Here is an example of two separate schemas in MySQL that are used for the Master and Work repository tables. SNPM1 holds the Master repository tables and SNPW1 holds the Work repository tables:

pic4


pic5


pic6

Here is an example of some of the type of information that is held in a typical repository table, in this case the name of a project that has been created in the Designer Navigator:

pic7


I hope that you have learned a little bit about our outstanding Oracle Data Integrator. I hope to meet you in one of my upcoming 11g or 12c classes!


Take Oracle Data Integrator Training


About the Author:

brentdayley

Mr. Brent Dayley is a Senior Instructor with Oracle University. Brent has worked at Oracle for over 10 years and teaches the SQL, PL/SQL, Application Express, Oracle Database Administration, SQL Tuning , MySQL and Data Integrator classes.

Comments:

Post a Comment:
Comments are closed for this entry.
About

Expert trainers from Oracle University share tips and tricks and answer questions that come up in a classroom.

Search

Archives
« May 2015
SunMonTueWedThuFriSat
     
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
      
Today