4. The migration project

4.1 Business drivers

The main business drivers for the migration project were:

• The SAS version 6.12 sunset
  The most important business driver for the migration project was the announcement by the SAS Institute to discontinue support for version 6.12 in the near future. It was decided that a business critical application should run in a supported environment.
• The number of users
  Because the migration project had become inevitable, NOTOX could benefit from moving from PC based SAS to server based SAS. NOTOX is a healthy, growing organisation and consequently the number of users has increased to a point where it becomes more attractive to move to server based SAS. This will decrease both licensing and maintenance costs.
• To capitalise on the benefits of new technologies
  Migrating would allow NOTOX to implement new technologies making the application more efficient, improve the look and feel and make it easier to maintain.

4.2 Project goals

The following project goals were defined:

• Migrate the FRS and HDS to a SAS 9.1.3 environment
  The main project goal is to migrate the FRS and HDS applications from version 6.12 to version 9.1.3 of the SAS system.
• Integrate both applications into one system and incorporate a third module to facilitate support
  To have one system incorporating three subsystems is more efficient in terms of validation, support and maintenance.
• Move from running the SAS System on desktops to run the SAS System on a server
  As the number of users is growing it has become more efficient to have to the SAS System server based.

4.3 Project approach

The project was executed in stages using a waterfall methodology. The approach is set out below.

4.3.1 Scoping study

First a scoping study was conducted to get a good overview the entire project. The following tasks were undertaken during the scoping study:

• Confirm project goals;
• Define project deliverables;
• Investigate migration and technology options;
• Perform a quick scan to determine the areas of the applications that need specific attention;
• Create a first cut estimate on effort and duration;
• Develop a detailed project plan.

4.3.2 Project definition

With the help of the scoping study document OCS Consulting and NOTOX selected the options which were deemed best fit. The options selected were:

• Migrate the user interface from SAS/AF to a Windows based user interface to be developed using Delphi.
  Because the move from PC based SAS to Server based SAS did not allow the use of SAS/AF as before, Delphi was selected to create the user interface.
• Use SAS Integration Technologies to communicate with SAS on the server.
  SAS Integration Technologies was selected as the middleware to communicate between SAS and the user interface. Having made the above selection, the project definition document could now describe exactly which areas of the FRS and the HDS will be affected and what the impact is. The document further contains a detailed project plan and detailed estimates on effort and duration.

4.3.3 Application design

As this project is a migration project, the original applications with corresponding documentation already existed. Therefore it was decided to create an application design document which combined both functional and technical specifications. The document is split up into two sections; one section details the programs on the SAS server and the other section describes the user interface.

The programs on the SAS server.
This section describing the programs on the SAS server mainly consists of high level specifications as most programs only needed to be migrated using base SAS. The SAS programs interacting with the new user interface needed to be created and were therefore described in full detail.

The user interface
This section describes all types of screen elements used in the application and how to handle these elements and their events. The screen elements are controlled by SAS through an information dataset containing the settings and values of each screen element. This allows control from within SAS on the values presented and the ability to hide or disable certain screen elements. Further to that it contains detailed flow diagrams of the application screens and their events. For each screen an example image of the screen to be created is displayed; the accompanying SAS information dataset to be expected and the events the screen has to be able to handle are provided as well.

4.3.4 Application development

The development work was split up into 4 main tasks which were executed parallel:

• Construction of the communication module to communicate between SAS and the user inteface;
• Development of the user interface;
• Creation of the SAS programs controlling the user interface;

• Migration of the SAS programs required by the system.

4.3.5 Application test

To test the migrated system a number of test scenarios were executed on the migrated FRS system and on the original FRS and HDS application. The results were subsequently compared.

4.3.6 Validation

The validation of the migrated FRS system is split up into 3 areas:

• Installation Qualification (IQ);
  When the FRS system was delivered to NOTOX an IQ was performed as specified in the FRS system installation guide. This IQ covered the installation of the SAS system, the import/transform procedure, the FRS system on the SAS server and the FRS system on the client PC. An IQ report stating that the installation was successful has been delivered to NOTOX.
• Operation Qualification (OQ);
  After the delivery was completed NOTOX performed an operation qualification procedure verifying that the FRS system is functioning as specified.
• Performance Qualification (PQ).
  When the OQ completed successfully a performance qualification was executed verifying that the system is functioning properly in a real life situation.

For the PQ the following scope was set:

• The transformed SAS datasets must be identical to the SAS datasets used by the previous version of the FRS and HDS for the project numbers used in the validation.
• The user interface must have equivalent or improved functionality compared to the previous version of the system.
• The user interface must offer the same or better information compared to the previous version of the system.
• The reports must be identical to the reports created by the previous version of the system or should differ from the previous version as expected due to approved change.