1- Library
Repository is used to store conversion tasks, the conversion task user through a graphical interface created can be stored in the repository.
Repository allows multiple users to share a conversion task, the task is to convert a folder in the form of group management repository, users can customize the folder name.
Library comes in two forms:
a, Kettledatabase repository, that is stored in a variety of common database repository type, user access to the repository resources by username / password, the default username / password is admin / admin and guest / guest.
Two, Kettle filerepository, preservation repository type in the folder server hard disk file, this type of repository without requiring the user to log in, direct manipulation.
The following figure shows the way to create a repository
2- Conversion
Conversion step, as will be appreciated the assembly of one or more different data sources into a data line. Then the final output to a certain place, such as a file or database
3- jobs
Job, you can schedule designed conversion, you can also perform some file processing (compare, delete, etc.), but also ftp upload and download files, send messages, execute shell commands, etc.
3.1 Conversion connection
Mainly shows the flow of data. From the input conversion operation such as filtration, to the output
3.2 connection work
To set the execution conditions:
1) the unconditional implementation
2) executed when the execution result is true a Job
3) When a Job execution result is false when the execution
4 cable
Connected or connected converting step Job (actually execution order) connection
5 Chef
It is a graphical user interface, using the SWT development, to design a job, conversion, SQL, FTP, e-mail, checklists, check the file exists, the script execute SHELL
6 Kitchen
Job execution engine, used for conversion, validation, FTP upload. You can perform tasks defined in xml format and stored on the database
7 Spoon
Kettle Spoon is another graphical user interface, the design for the data conversion process
8 Pan
Pan is a data conversion engine that is responsible for reading and writing different data sources and data conversion.
pan.sh -file="/PRD/Customer Dimension.ktr" -level=Minimal