Spoon allows you to store your transformations and jobs under two different configurations: file-based and database repository. In contrast to the file-based configuration that keeps the transformations and jobs in XML format as *.ktr and *.kjb files in the local filesystem, the database repository configuration keeps the same information in tables in a relational database.
Although working with the file-based system is simple and practical, the database repository method can be convenient in some situations.
The following is a list of some of the distinctive repository features:
Repositories implement security. In order to work with a repository, you need credentials.
Repositories are, by their nature, prepared for basic team development. The elements you create (transformations, jobs, database connections, and so on) are shared by all the repository users as soon as you create them.
The Enterprise Repository is a Java content repository capable of more robust and scalable collaborative functions such as version control, locking, and more.
Before you decide on working with a repository, you have to be aware of the file-based system benefits you lose. Here are some examples:
When working with the database repository-based system, you need access to the repository database. If for some reason you cannot access it (for example, network problems), you will not be able to work. You don't have this restriction when working with files, where you only need the software and the transformation and job files, that is, the
.ktrand.kjbfiles.When working with the database repositories, it is difficult to keep track of the changes. Working with the filesystem, it is easier to know which jobs or transformations were modified. If you use Subversion or Git, you even have a control version that allows you to examine the history of changes and to recover older versions of your work if necessary.
Suppose that you want to search and replace some text in all the jobs and transformations. If you are working with repositories, you would have to do it for each table in the repository database. Whereas working with the file-based system, this task is quite simple. For example, you could create a Sublime project - available for downloading at www.sublimetext.com - open the root directory of your jobs and transformations, and do the task by using the Sublime utilities.
As explained in Chapter 1, Getting Started with Pentaho Data Integration, there is a third method, File repository, that is a mix of the two mentioned earlier. It's a repository of jobs and transformations stored in the filesystem.
Note
The use of the File repository is similar to the database repository. Therefore, we will not explain it in this appendix. You should not have any difficulty in trying it once you understand how to work with the database repository.
This appendix shows you how to create a database repository and how to work with it. You can try repositories and decide for yourself which method, database repository-based or file-based, suits you best.