Oracle Projects Migration/ Data Conversion


During my work on a projects conversion, I came accross this very nice document. This document is very informative and so sharing with you all. Happy reading!

Main Article:

In this article I will be explaining the general steps involved in any Conversion/Data Migration of Oracle Projects module.

At the end of this article, you would have learned:

  • Stages in Oracle Projects Conversion.
  • How to setup the Oracle Projects module for the conversion/Data Migration.
  • Options for the Load (flat file, csv, or direct Loads).
  • Oracle Projects AMG APIs needed to perform the Conversion.
  • Testing the Conversion Process.
  • Verifying the Conversion Process.

Scenario:

Company ‘XYZ’ is using a Project Management and Accounting Software for years long. The Management has decided to move from their existing system to Oracle Projects module because of its vast functionality and integration with other financial modules.

How to deal with it?

Now the question arises: What data to migrate from the legacy system to Oracle Projects?. Well, it depends upon the type of projects.

If the Projects are used for Internal Administration and tracking of costs, you may want to migrate the existing projects, tasks(the work break down structure), Cost Budgets, Cost (Timecards, Employee Expenses, Miscellaneous Expenses) etc.

If the Projects are used for billing the Clients for the work done (Typical Contract Projects), then you may want to Revenue, Agreements (Contracts), Revenue budgets and Invoices in addition to the above data.

Once the decision is made to which data to migrate, then the next step would be setting up the Oracle Projects for the conversion purpose, which we will see in detail sooner. Once the System has been setup, the technical elements (programs, concurrent processes etc) have to be created in order to migrate the data from Legacy System to Oracle Projects.

Stages in Oracle Projects Conversion

1.     The First Stage is to obtain the data from the Legacy System which needs to be migrated to Oracle Projects.

2.     The Second Stage will be most crucial step in the process which is to massage the data according to the Oracle Projects Conversion Interface (Programs built using AMG APIs). This Step is indeed time consuming, manual labor intensive to massage and rectify the errors etc. But completing this step successfully pays dividends in the consecutive processes / Stages.

3.     The Third Stage is uploading the data obtained from legacy systems into the Staging Area (Staging Tables created to hold the data temporarily till it gets migrated into Oracle Projects). Once the data is uploaded to the Staging tables, the programs built for migration (We will see how to build these programs in detail) will validate the Staging Table data to confirm that it is in compliance with the Projects Conversion Program (The AMG APIs used in the programs indeed needs data in certain format, also the data should be validated against the Oracle Projects Setup. For instance, when migrating the cost or hours from legacy to Projects, we might need to validate if the expenditure type is already setup in Oracle Projects, if the expenditure type is not setup, the program/APIs will throw an error. So it is always better to capture these kind of scenarios in the Validation Step of the Migration.

The Second Stage and Third Stages are repetitive until you get the valid data from the legacy system which can be migrated into Oracle Projects without any errors or issues.

The Fourth Stage is the actual migration process which will migrate the data from the Staging Tables to the Oracle Projects Base tables. Once this step is done, the projects, tasks and other data are available in Oracle Projects for use.

Before going through the stages, we will look at some of the basic setups that need to be done in Oracle Projects.

Oracle Projects Setup For Conversion

Product Code:

The Product Code needs to be setup in Oracle Projects in the AMG Gateway – Source Products Form in the Oracle Projects Implementation Super User Responsibility. This setup is mandatory since this product code needs to be passed when using the Oracle Projects AMG APIs

Project Types and Project Templates:

The project types and project templates for conversion projects need to be setup up. This is a mandatory setup since while migrating projects we need to tell the APIs which project template/type the projects use.

For Contract Projects, setup the Contract Project Type Template. For administrative or internal projects, setup the Indirect Project Type templates.

If you are migrating Cost and Revenue Budgets, then the Plan Types need to be attached to the templates in order to create the budgets for the migrated projects.

Implementation Option Setup:

Project Numbering: This implementation option is by default set to ‘Automatic’ which means when creating projects in Oracle Projects, the project number is automatically derived and users are not required to provide any project numbers. This option is best suitable when creating projects in Oracle Projects. But when migrating the projects from the third party systems, there is an option to migrate the projects with the same project number as in the legacy system. This is not mandatory but is recommended since it will be easy to refer back the projects in the source system using the project numbers.

In order to pass the project number to the Migration program, this implementation option needs to be setup to ‘Manual’. Once the migration is done, this setup can be reverted back to ‘Automatic’.

Setup Transaction Source:

The Transaction Source needs to be setup in Oracle Projects in the Transaction Sources form in Oracle Projects Implementation Super User Responsibility. This is a mandatory setup for the Costs/hours migration from the legacy system to Oracle Projects. We need to tell the migration API’s what the source system is and how the data is handled when it is imported to Oracle Projects.

Setup Expenditure Types:

Expenditure Types are needed to categorize the cost/hours when it is imported to Oracle Projects. This is a mandatory setup for Cost/hours migration. We need to tell the system which expenditure type the cost/hour belongs to.

Setup Employee Cost Rates:

Setting up cost rates for employees is not mandatory. But if you need to cost the hours that are migrated in the system, the labor cost distribution process in Oracle Projects do need the rates setup in order to calculate the costs.

But if you are migrating the costs directly from the legacy instead of hours then this step is not needed. But ideally the cost rates are required in a general production scenario wherein the employees/contractors enter their timecards.

You can setup job rate schedule, employee level rate schedule or employee level overrides. Alternatively, the costing client extension can be setup to calculate the cost according to the business scenario.

Refer to the Oracle Projects User Guide for how to setup the employee cost rates.

First Stage: Obtain Data from Legacy System

The first stage deals with obtaining the data from the legacy system in the desired format. The data can be obtained in the form of flat text file or comma separated file csv, tab delimited file or file with any delimiters. Generally tab delimited files are recommended since comma separated files behave strange when there is a comma in the data itself.

If there is a database link created between the Source Legacy database and the Oracle Projects Database then the data can be obtained directly using the select statements against the Source DB from within the Oracle Projects DB. But this method is not preferred as it is more performance intensive when it comes to selecting large data over the network.

For Projects Migration, generally 2 files are obtained. One file for Projects Data and the other file for Tasks Data.

For Transaction Migration, single file is enough with all the cost/hours data.

For Cost/Revenue Budget migration, single file is enough with all the Budgets Data.

Create SQL Loader concurrent program which will upload the obtained data into the Oracle Staging Tables.

Also it is always the best practice to create a control table in the Staging area, which will control the data migration. For example your control table might look like the one below:

Parameter Type Parameter Parameter Value
Template Contract Contract_Template
Template Indirect Indirect_Template
Expenditure Type Hours Labor
Expenditure Type Expenses Employee_Expense
Transaction Transaction Source Legacy1
Product Code Product code LEGACY1
Project Publish Workplan Yes
Project Baseline Workplan Yes
Cost Budget Baseline Yes
Revenue Budget Baseline Yes

This control table is looked upon by the migration program. So whenever there is a change in the templates, expenditure types it is easy to change this control table instead of the code. So the advise is never hard code any values in the code, always handle it using the control table.

Also it will be better to have a form based on this table, so that this table data can be changed from the front end.

Second Stage & Third Stage: Validate and Format the Data

I am coupling the second and third stage because both are interdependent. Validating data is very important and it prevents some of the time consuming tasks in actual migration such as trouble shooting the errors due to the invalid data.

Below are some of the key validations that need to be done before doing the actual migration.

Projects/Tasks Migration:

Though the projects and tasks are in different staging tables, the migration of projects/tasks is doing using a single program. We can always migrate projects and tasks separately, but the issue is with the performance when adding task by task to each project. So it always better to create projects and tasks together because of the bulk loading of tasks.

Project/Task – Setup Validations: 

  • Validate the Product code is setup.
  • Validate if the required Project Templates are setup.
  • Validate if the Project Numbering is set to ‘Manual’ for creating projects with the   predefined project numbers.

Project/Task Data Validations:

  • Validate if the project name is unique. Project with the same name should not exist in Oracle Projects.
  • Validate if the project number is unique. Project with the same number should not exist in Oracle Projects.
  • Validate if the project long name is unique. Project with the same long name should not exist in Oracle.
  • Validate the project reference(this field is mandatory in the projects file, it can be the projects identifier of the source project or project number of the source project, but it has to be unique in the source system as well. This field needs to be populated in all the converted projects in order to track back and identify the project in the source system)
  • Project name and project number should be 30 chars in length. Project long name should be 240 chars in length. Project Description should be 250 chars in length. Project description is not a mandatory field when creating project.
  • Check if the project has a project manager and the project manager is active in Oracle HR and has an assignment and a Job assigned. Also the project manager has to be active from the project start date, else you cannot create a project with that project manager.
  • In case of contract projects, check if the customer of the project is a valid customer defined and with a valid Bill To site assigned.

Apart from the above necessary validations, you may have to validate the additional data such as Projects DFF Data you may want to populate with your custom field values. For example you may want to populate the Project cost center value in the Segment1 of the Project DFF. In such case you have to validate if the cost center value is a valid value for that Segment1 (sometimes you may have attached an LOV to that segment1, so in that case, the cost center has to be validated against that LOV Values).

For tasks, values for task types, work type, task manager has to be validated. Task types and work types have to be defined in Oracle Projects before the task with those values are migrated, else the task will not be created.

Cost/hours validation

Setup Validations:

  • Validate if the Transaction source is setup.
  • Validate if the Expenditure type is setup.

Data Validations:

  • Check if the hours value is greater than zero.
  • Check if the employee number is valid in HR and is active on the timecard date.
  • If the transaction source is setup as costed, then the cost has provided while migrating the transactions. If the transaction source is setup as accounted, then the code combination ids need to be provided when migrating transactions.

Apart from the above validations, you may want to validate the additional DFF segments that you are going to populate for that expenditure item.

Budgets Validation

Setup Validations:

  • Validate the project template has the required financial plan type attached. Financial plans are the project management versions of the Budget types in the Forms applications.
  • Budget amount has to be greater than zero.
  • There is no need to create revenue budgets if the ‘Baseline funding without budget’ option is checked at the project or project type level. Whenever the funding is created for the contract project and is baselined, the revenue budget is automatically created and baselined. If that option is not checked, it is necessary that a revenue budget with the same amount as the funding amount needs to be created and baselined in order to baseline the funding.

Data Validations:

The cost budget for the project can be from the source system’s budgeting system. If there is no budgeting in the source system, a cost budget with the total cost of the project can be created in Oracle Projects.

For revenue budgets, it has to be equal to the funding amount of the project. If there is no funding amount in the source system, the sum of the revenue amount can be the funding amount and it is the revenue budget amount as well.

Agreements and Funding Validation

Data Validations:

  • Agreement type should be valid.
  • Agreement Amount should be greater than zero.
  • Hard Limits can be setup according to business rules. If the hard limits are setup for revenue and invoice then the revenue and invoice has to be within the funding limits for that project.
  • Funding amount has to be within the Agreement amount.
  • If the funding at the top task level, then the ‘Customer at top task’ has to be enabled and the customer should have been assigned at the top task.
  • Funding amount should be same as the Revenue budget amount which in general will be same as the total revenue amount for that project. If there are no hard limits then the revenue or invoice can exceed the funding amounts.

Records which fail the above validations have to be rectified before doing the actual migration.

Revenue and Invoice Validations

Data Validations:

  • Project / Task should already been converted to Oracle.
  • Event amount should be non zero.
  • For revenue event revenue amount should be populated.
  • For invoice event invoice amount should be populated.

Generally for a project, the total revenue is obtained from the source system and is created as a revenue event for that project. The total invoiced amount is calculated per project and an invoice event is created for each project.

Once these events are created successfully in the system, the Generate Draft Revenue process and Generate Draft Invoice process needs to be run so that the desired revenue and invoices are generated.

The revenue and invoice automatic approval and release client extensions can be used to automatically release the revenue when it is generated and approve/release invoices respectively.

If the revenue amounts are already interfaced to General Ledger (GL) through a different interface, then uncheck the ‘Interface Revenue to GL’ option in the implementation options and run the ‘Interface Revenue to GL’ process in Oracle projects. This will turn the flags in the revenue records as accepted in GL, though it is not interfaced. Once this is done, revert back the implementation option back to its original state.

If the invoice amounts are already interfaced to Accounts Receivables (AR) by different means, it is not desired to interface the projects invoices to AR again since it will double the invoice amount in AR. In this case, we do not have an implementation option like we had for Revenue. So a script can be created to update the Invoice’s flag to Accepted State. Alternatively the generated projects invoices can be interfaced to AR, tied back to Oracle and then the invoices can be deleted in AR.

Stage 4: Actual Migration

Once the data is validated, the program for conversion is executed to migrate the data into oracle projects base tables. There might be still errors due to AMG APIs which has to analyzed and resolved. But the chances of such AMG API issues are just below 10% in any migration (based on my experience in Oracle Projects Conversion).

Below is a table with Conversion and which AMG APIs are used for that conversion:

Conversion

AMG APIs

Projects/Tasks Conversion PA_PROJECT_PUB.CREATE_PROJECT
Budgets Conversion PA_BUDGET_PU B.CREATE_DRAFT_BUDGET, PA_BUDGET_PUB.BASELINE_BUDGET
Agreements PA_AGREEMENT_PUB.CREATE_AGREEMENT
Funding PA_AGREEMENT_PUB.ADD_FUNDING
Revenue/Invoice PA_EVENT_PUB.CREATE_BILLING_EVENT
User Defined Attributes (UDA) PA_PROJECT_PUB.LOAD_EXTENSIBLE_ATTRIBUTE

For Transactions (cost/hours) migration, there is no APIs to create the expenditures in Oracle. The pa_transaction_interface_all table needs to be populated with the migration data and once it is populated, the PRC: Transaction Import process with the Transaction source as parameter needs to be run in Oracle Projects. All invalid records need to be rectified in order to migrate all the transactions.

The rejected records can be found in the same interface table with the transfer_status_code as ‘R’.

Conversion Tips:

1.     Make sure the templates are defined properly and exactly the way it is needed. Once the projects are created using the templates and the template was wrongly defined, then it takes ages to rectify the converted projects.

2.     Create the conversion program to operate in two modes: Validate, Run. A concurrent process with a parameter called mode accepting Validate/Run can be created. So the same concurrent program can be used to validate as well as run the actual migration.

3.     It is a good practice to have source Project id / Project Number as parameter to the projects conversion program. This will allow us to test the conversion for a single project and validate the data for that project.

4.     The validation process can write the invalid records to the output file. So once validation process completes, the output will have all the invalid records which needs to be rectified.

5.     Create a separate concurrent program to know the status of the already running migration process. If you want to know where the migration process is in terms of the number of records migrated, number of records rejected etc. If the volume of the migration data is huge, then it is likely possible that the conversion programs may run for hours. So in these scenarios this concurrent program can be helpful in finding the status of that migration process.

6.     For Transactions migration, the custom program written to populate the interface table can kick off the PRC: Transaction Import process and wait for its completion. Once the transaction import completes, the custom process can print the invalid records from the interface table to the output file.

7.     There are APIs to publish and baseline the workplans created as a part of projects migration. But these APIs need to be used with care. There are lot of performance issues and bugs when using these APIs.

Original Post

Author: Sathish Raju

Website: http://www.projectsaccounting.com

Advertisements

AR Invoice Interface


AR Invoice Interface

The main three steps for AR Invoice Interface are:

1] Put the data into your staging tables.

2] Calls your package to validate the data and load into AR Interface tables (RA_INTERFACE_LINES_ALL & RA_INTERFACE_DISTRIBUTIONS_ALL).

3] Then submits a concurrent request for AutoInvoice.

If any errors occur it can be found in ra_interface_errors_all table. The concurrent program has 2 stages. First the Master program fires which intern kicks of the Import Program. Once this is completed data is inserted into the following tables.

1) ra_customer_trx_all (Invoice Header Info)

2) ra_customer_trx_lines_all (Invoice Line Level Info)

3) ra_cust_trx_line_gl_dist_all (Accounting Info. One record for each Account Type is inserted into this… ex. Receivable Revenue Tax Freight etc)

4) ar_payment_schedules_all (All Payment related info)

Validations:

Validation are generally done on the below columns.

  • Batch_source_name
  • Set_of_books_id
  • Orig_sys_batch_name
  • orig_system_bill_customer_ref
  • orig_system_bill_address_ref
  • Line_Type
  • Currency_Code
  • Term_name
  • Transaction_type
  • Interface_line_attribute1-7
  • Account_class
  • Accounting Flexfields segments

1- AR Transaction Type Validation: Check if the Transaction type provided in data file is defined in AR transaction types (RA_CUST_TRX_TYPES_ALL)

2- Transaction Batch Source Validation: Check if the source provided in data file is defined in AR transaction Batch source (RA_BATCH_SOURCES_ALL).

3- Invoice Currency Validation: Check if the currency provided in data file is defined in AR Currency (FND_CURRENCIES).

4- Customer Validation: Check if the Bill to Customer Number, Ship to Customer Number, Bill to Custom Location, Ship to Customer Location provided in the data file is defined in AR Customer (RA_CUSTOMERS).

5- Primary Sales Representative Validation: Sales representative number to be hardcode to “-3” for “No Sales Credit.”

6- Term Name: Check if the Term name provided in the data file is defined in Payment terms (RA_TERMS)

7- Inventory Item Validation: Check if the Item provided in data file is defined in Inventory Items (MTL_SYSTEM_ITEMS).

8- Unit of Measurement validation: Check if the UOM provided is defined in MTL_UNITS_OF_MEASURE Table

9- Invoice Tax Code Validation: Check if the Tax Code provided in data file is defined in AR_VAT_TAX_ALL_B Table.

10- Invoice GL Date Validation: Check if the GL Data of provided invoices is in open period.

For MOAC:

You need to add the below columns and need to do validations if your application supports MOAC.

  • conversion_type
  • conversion_rate
  • conversion_date

Sample Code to run Autoinvoice Master Program:

DECLARE
v_phase 	VARCHAR2(100);
v_dev_phase 	VARCHAR2(100);
v_status 	VARCHAR2(100);
v_dev_status 	VARCHAR2(100);
v_message 	VARCHAR2(100);
v_reqid 	NUMBER(15);
v_pid 		BOOLEAN;
v_user_id  	NUMBER(30);
v_batch_source_id NUMBER;
v_order    	NUMBER;
v_org_id   	NUMBER;
v_resp_id  	number;
v_resp_appl_id 	number;
v_appl_short_name fnd_application.application_short_name%TYPE;

CURSOR c1 IS
select fcr.responsibility_id
,fr.application_id
from fnd_concurrent_requests fcr
,fnd_responsibility fr
where fcr.request_id = '${4}'
and   fcr.responsibility_id = fr.responsibility_id;

CURSOR c2 IS
select fa.application_short_name
from fnd_concurrent_programs fcp, fnd_application fa
where fcp.concurrent_program_name = v_program_short_name
and fcp.application_id = fa.application_id;

CURSOR c_batch_id IS
SELECT 1, batch_source_id, name
FROM apps.ra_batch_sources_all
WHERE name IN (SELECT distinct a.batch_source_name
FROM xxfin.xxfin_ar_ol_invoices a
WHERE a.batch_source_name like '%DEBIT'
AND filename = '${file1}')
UNION
SELECT 2, batch_source_id, name
FROM apps.ra_batch_sources_all
WHERE name IN (SELECT distinct a.batch_source_name
FROM xxfin.xxfin_ar_ol_invoices a
WHERE a.batch_source_name like '%CREDIT'
AND filename = '${file1}')
order by 1;

BEGIN
open c1;
fetch c1 into v_resp_id,v_resp_appl_id;
close c1;

open c2;
fetch c2 into v_appl_short_name;
close c2;

FOR v_batch_data IN c_batch_id LOOP

fnd_global.apps_initialize('${FCP_USERID}',v_resp_id,v_resp_appl_id);

v_reqid := fnd_request.submit_request('AR',
'RAXMTR',
NULL,
to_char(trunc(sysdate),'YYYY/MM/DD HH24:MI:SS'),
FALSE,
'1',
-99,
v_batch_data.batch_source_id,
v_batch_data.name,
to_char(trunc(sysdate),'YYYY/MM/DD HH24:MI:SS'),
NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,
NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,
'Y',
NULL);
commit;
v_pid := fnd_concurrent.wait_for_request(v_reqid,
3,
0,
v_phase,
v_status,
v_dev_phase,
v_dev_status,
v_message);
END LOOP;
END;

GL Budget Interface


GL Budget Interface

This program lets you prepare and analyze your budget outside of General Ledger and then transfer your budget information into General Ledger. This enables you to perform your budgeting in the environment you choose, and still maintain the integrity of your database.

Interface Table:

GL_BUDGET_INTERFACE is used to upload budget data into your Oracle General Ledger application from an external source. Each row includes one fiscal year’s worth of budget amounts for an account. When you load this table, you must supply all NOT NULL columns with data. In addition, you must supply a valid account combination in the SEGMENT columns appropriate to your chart of accounts. Finally, you must supply the budget amounts in the appropriate AMOUNT columns.

The mandatory (not null) columns of the Interface table are:

  • BUDGET_NAME
  • BUDGET_ENTITY_NAME (the budget organization)
  • CURRENCY_CODE
  • FISCAL_YEAR
  • UPDATE_LOGIC_TYPE (A for Add, R for Replace)

Other important columns are:

  • SET_OF_BOOKS_ID
  • BUDGET_VERSION_ID
  • PERIOD1_AMOUNT through PERIOD60_AMOUNT
  • SEGMENT1 through SEGMENT30

Validations:

  • Budget Name and Budget Entity Name
  • Currency Code
  • Account Segments(Code Combination in GL_CODE_COMBINATIONS Table)

To upload a budget:

  1. Navigate to the Upload Budget window (Budgets > Enter > Upload).
  2. Enter the Budget and Budget Organization.
  3. click Upload.

General Ledger submits a concurrent process to upload budget information from the GL_BUDGET_INTERFACE table.

Budget Upload Validation:

Budget Upload validates all of your budget information for compatibility with General Ledger. Budget Upload also checks to make sure that the following conditions are true:

• Your account is assigned to a budget organization

• The budget entry type for your account is Entered

• Your budget is not Frozen

• Your budget organization is not Frozen

• Your budget fiscal year is open for your budget

Once updated, General Ledger automatically deletes the rows of budget records in the Budget Interface table.

Base Tables:

  • GL_BUDGETS
  • GL_BUDGET_ASSIGNMENTS
  • GL_BUDGET_TYPES

For more information see Oracle General Ledger User Guide

Data Migration vs. Data conversion


Data Migration vs. Data conversion

When we need to enter data into oracle Apps, following are the few techniques:

• The Data can be entered using the application Screens (for small amount of data, like creating PO, entering sales orders using Oracle Apps screens).
• The data can be entered using Oracle’s Open System Interface (for regular operations e.g. for moving data from one module to another).
• The data can be stored in the database table directly (Not recommended by oracle and can be very risky, because on any event data is going to be stored in many tables and data should be validated before inserting into tables that may cause data integrity and inconsistency problem, sometimes it may corrupt the data completely.).
• Using third party tools like data loader (It is also can be used when data is relatively small (25-200 records) because it captures the keystrokes and works like manually entering the data into Oracle form but much faster as process is automated).

What is the need of Migration/Conversion?

Migration/Conversion are required when we are upgrading to one version to another (e.g. Oracle Apps 11.5.7 to Oracle 11.5.10) or moving data from some legacy system to Oracle Apps. There will be bulk  of data (sometimes millions or even more than that) that needs to be moved from one system to another  and  before moving the data it should be validated and only valid records should be entered into Oracle Apps.

If both the systems (Target and source) are not having same structure for data (Tables are not same/Table Structure is not same/The data is being stored in database is not same), it needs to be translated (e.g. upgrading from Oracle 11i to R12 where table structures are not same) then we say it as conversion (any kind of translation of data on Source data to make it suitable for Target system) otherwise migration.

 

What is Migration?

 

Migration of data means moving the data from one system to another using Interface Programs/APIs where both the systems have same structure of data.

Process of Migrating of data:
• Identify the data to be imported to new system (Business requirement).
• Extract the data into flat file/Staging table
• Load the data into Interface Table(using SQL* Loader/DB Link/Others) after validation(If loading the data using Interface)

What is Conversion?

Conversion of data means translating the data to suite target system (data should be formatted according to target system )  and then move the translated data using Interface Programs/APIs.
• Identify the data to be imported to new system (Business requirement).
• Extract into flat file/Staging table
• Translate/Convert/Format the data
• Load the data into Interface Table(using SQL* Loader/DB Link/Others) after validation(If loading the data using Interface) and then launch standard Interface concurrent program to load the data to Oracle Apps Base Tables
• If using API, fetch the data, validate it and then call API to import the data

 

How conversion/Migration and interface differ?

 

There are good numbers of parameter on which they can be categorized. Take few of them:
Frequency
• Conversions/Migration are a one time event
• interfaces are ongoing
Occurrence in the project timeline
• conversions/Migration executed before production
• interfaces executed during production
Manner of execution
• Conversions/Migration are batch
• Interfaces may be batch or real time
Complexity
• Conversion/Migration does have very complex, it’s totally depends upon the data mapping activity.
• Coordinating with other systems make interfaces more complex
Maintenance
• Maintenance of interface is bit cost intensive task.

 

ABOUT INTERFACES


 

ABOUT INTERFACES:

In Oracle Apps Interfaces are generally tables, which act as a medium to transfer the data from one module to another module or to transfer the data from legacy system into Oracle Applications. There are 352 tables provided by the Oracle Package. Each module has its own Interface Tables.

A typical path to transfer the data from Legacy System to Oracle Apps:

What is Interfacing?

It is the process of converting the records from one format to another format. The main components of this interfacing are

• Transfer Program

• Interface Table and

• Import Program

A] Transfer Program:

If the source modules data are implemented in Oracle Applications then the Transfer Programs are integrated with the Package. If the source modules are implemented in external system (i.e. other than Oracle Applications) then we have to develop our own Transfer Programs. Generally these Transfer Programs are developed using PL/SQL, JAVA or SQL Loader.

What they do?

  • It maps the columns of source table with the columns of Interface Tables.
  • It performs Row Level and Column Level validations.
  • It transfers the data from Source to the Interface Table.

B] Interface Tables:

The Interface tables basically have 4 types of columns.

  1. Mandatory Columns.
  2. Conditionally Required Columns.
  3. Optional Columns.
  4. Internal Processing Columns.

Mandatory Columns:

These are the main columns which are required in the destination tables (i.e. Oracle Application Module Tables). With the help of mandatory columns only the Import Program will converts the records from source to destination.

Conditionally Required Columns:

The values for these columns are based on the values of Mandatory columns. For Example: If you are converting foreign currency transactions to INR then it as compulsory to provide conditionally required columns like Currency conversion rate, Conversion Time and Conversion Date.

Optional Columns:

These are used when a client wanted to transfer some additional information from source to destination. These are based on client’s requirement.

Internal Processing Columns:

Status and Error Message columns are called Internal Processing Columns. These are specific only to Interface Table. These columns are going to be used by the Import Program to update the status and error message, if the record fails its validation while importing from Interface Table to the Destination Table.

C] Import Program:

For all Interface Tables, Oracle Application Package is going to provide Import Programs. These are generally registered with destination modules. These Import Programs are designed using PL/SQL, JAVA, C, C++, etc.

What they do?

  • It maps the columns of the Interface Table with one or more columns in the destination table.
  • It performs row level and column level validation.
  • It imports the data from Interface Table to the Destination tables, if the records validated successfully.
  • It deletes all the successfully validated records from Interface Table.
  • If the record fails its validation then the Import Program will update the status and error message columns of Interface Table.

Interface Vs. Application Program Interface (API):

Interfaces are used to transfer the data from legacy system to Oracle Application system where as API is used to convert the data from one form to another form with in the Oracle Application Module.

Previous Post:

 

Interfaces in Oracle Application: An Introduction


Interfaces in Oracle Application: An Introduction

 

What are Interfaces?

  • Interfaces are used in Oracle Applications to integrate external systems and Data Conversion.
  • The interfaces are mainly used to either transfer data from Oracle Applications to a flat file or data from legacy system to Oracle Applications.
  • Used extensively at the time of Data Conversion from legacy/ old systems to a fresh implementation of Oracle Applications.
  • Used also at regular intervals when data transfer is from other live systems if the systems are not defined in Oracle Applications implementation.
  • Oracle provides flexible and flexible tools in the form of Interface programs to import the master and transactional data like Customers, Invoices, and Sales Orders etc from external systems into Oracle Applications.

Types of Interfaces

There are two major types of Interfaces:

  • Inbound Interface : These interfaces are used to transfer data from external systems to Oracle Applications.
  • Outbound Interface :  These interfaces are used to transfer data from Oracle Applications to external systems.

Two other distinctions of Interfaces:

  • Open Interface: If the interface logic is provided by Oracle Applications, it is called an Open Interface.
  • Custom Interface: If the interface logic needs to be developed by the implementation team, it is called a Custom Interface.

Interface Components

Open Interface Logic

  • First the data from the source application is loaded into a database table (called Interface table).
  • Then the provided validation program logic validates the records whether they are correct or not .
  • If the validation fails, the errors are transferred into another table (called Error Table).
  • If the validation succeeds, the correct records are transferred through a process into the destination application table.

Components of an Interface

a] Source Application:

You obtain data from a source application to pass on to a destination application for further processing and/or storage.

b] Source Data Issues:

Type of file, Size, Frequency of upload, Record Length (Variable or fixed), Delimiter, Datatype for each field, Any unwanted data, Naming convention and uniqueness of file, Location of the file, Access on the file.

c] Destination Application:

You send data to a destination application so that the application can perform further processing and/or storage.

d] Interface Table:

For inbound interfaces, the interface table is the intermediary table where the data from your source application temporarily resides until it is validated and processed into the destination application.

e] Identifier columns:

Uniquely identify rows in the interface table provide foreign key reference to both the source and destination applications.

f] Control Columns:

  • Control columns track the status of each row in the interface table, as it is inserted, validated, rejected, processed, and ultimately deleted.
  • WHO columns are also control columns.

g] Data Columns:

  • Stores the data that is being converted.
  • Required columns store the minimum information needed by the destination application to successfully process the interface row.

h] Derived Columns:

Derived columns are created by the destination application from information in the required columns.

i] Optional Columns:

Optional columns are not necessarily required by the destination application, but can be used by the destination application for additional value-added functionality beyond the basics.

j] Error Table:

  • For inbound interfaces, the errors table stores all errors found by the validation and processing functions.
  • In some cases, the errors table is a child of the interface table. This allows each row in the interface table to have many errors, so that you can easily manage multiple errors at once.
  • In other cases, the errors are stored in a column within the interface table, which requires you to fix each error independently.

Developing an Interface

1] Identification:

Find out if there exists an Open Interface to carry out the functionality.

2] Creation of Pre-Interface table ( staging Table):

A table in the format of the data file which can be pruned to load as clean a data into the Interface table.

3] Load data into Pre-Interface table:

SQL*LOADER can be used to load the flat file into the pre-interface table.

4] Validate data in the Pre-Interface table:

Basic validation of the data loaded into the Pre-Interface table can be carried out like:

  • For checking NULL values in required columns
  • Checking for Foreign Key and Quick Code values.
  • Duplication Validation
  • Business Rule validation

5] Mapping the values:

Generated fields in Oracle Applications can be mapped in this step to either default values or sequences.

6] Load data into Interface table:

  • Once the data is as clean as you can get it, the data can be inserted into the Interface table.
  • At such a time, certain columns, which are necessary in Applications but not found in legacy system, need to be populated accordingly like WHO columns.

7] Run the interface program

8] Check for Errors

9] Report on the Interface