Designing Your Source Extractor

You must consider many issues when designing your source extractor. For example, ideally your source extractor will extract a reasonable amount of data overnight. Also, if you have international users, the source extractor should be easy to localize and should include appropriate code pages. You should also consider these questions:

• What level of sophistication is required to install and configure the source extractor?
• Does a user-driven source extractor make sense?
• Where are the files going to be edited after export and before import?
• How will the files be transferred from the extract location to the import location?
• Will file transfer modify the files?
• Is the existing system going to be in use?

Also consider the following additional issues when creating your source extractor:

Existing data   If possible, copy, don't move, the data from your existing system, so that you will have an additional copy of the data.

Group migration   If your existing system has account information in the directory, select accounts to migrate by department or location.

Permissions   Administrator or user authentication should be required to copy data from the source system.

Primary mailbox files   Mailbox files need to be edited if you're changing the directory name. It is best to put all the directory entries into one primary file.

Mailbox names   You cannot change the common or directory name of a mailbox after you create it. If you have a new naming convention, modify every primary file that contains mailbox creation information, or modify the source extractor to select the new directory names from a file.

You should also consider each of the following issues:

• Choosing the addressing conventions
• Choosing information to extract
• Configuring command-line options
• Configuring special features