Installing JD Edwards EnterpriseOne on Amazon RDS for Oracle

Marc Teichtahl is a solutions architect at Amazon Web Services.

Amazon Relational Database Service (Amazon RDS) is a web service that makes it easier to set up, operate, and scale a relational database in the cloud. It provides cost-efficient, resizable capacity for an industry-standard relational database and manages common database administration tasks. Amazon RDS offers all the benefits of high availability, security, and scalability, while eliminating the need to manage infrastructure, upgrades and updates and the complexity of handling a highly available environment.

Recently, we have been asked by many enterprises to provide guidance on how to install Oracle JD Edwards EnterpriseOne on Amazon RDS. This blog post provides an overview of installation, highlighting the modifications required to ensure success. Accompanying this post is a detailed whitepaper, providing a step-by-step guide for this process.

Key elements of the installation process for JD Edwards EnterpriseOne on Amazon RDS for Oracle include creation of an Amazon RDS Oracle instance, configuration of the SQL Plus client, installation of the platform pack, and modification of the original installation scripts. Much of this process is straightforward. However, significant changes to the original installation scripts are required. This blog post describes those changes.

You can follow the standard JD Edwards installation process found in the JD Edwards EnterpriseOne Applications Installation Guide until section 6.8, “Understanding the Oracle Installation.” At this point, the installation process varies when using Amazon RDS for Oracle.

The standard installation process creates a series of installation scripts. However, the process assumes these scripts will be executed on the database server. In this case, we are targeting Amazon RDS, and we can’t execute installation scripts on the Amazon RDS instance itself.

For the functionality we require, we need to modify the standard installation scripts in four key ways:

1. Move the data pump files from various directories on the deployment server installation media to the DATA_PUMP_DIR directory on the RDS instance using DBMS_FILE_TRANSFER.PUT_FILE. For more information, see the Amazon RDS documentation.
2. Amazon RDS only supports Oracle Managed Files (OMF) for data files, log files, and control files. Thus, when creating data files and log files, you can’t specify physical file names (for more information, see the Amazon RDS documentation).
   Make syntax changes to the “CREATE TABLESPACE” statements accordingly. Take the following example.
   
   This code must be modified as shown following.
   
3. Change the name of the pristine data dump file and also the import data script for the TEST environment and pristine environment. The standard scripts change the import directory; we are going to change the file name. Make these changes to ORCL\InstallOracleDatabase.BAT, as shown following.
   
   
4. Change the database grants to remove “create any directory” because this grant doesn’t work on Amazon RDS for Oracle. Amazon RDS for Oracle doesn’t support creating directories on the RDS instance.
   For example, take the following Create_dir.sql file statement.
   
   Here, you change this statement to the statement following.
   

The preceding guidelines should give you enough insight into adapting JD Edwards EnterpriseOne installation to take advantage of the benefits of Amazon RDS for Oracle. For more granular step-by-step instructions, see the whitepaper.