Just throwing up an idea... I don't have the time in short notice to do
anything on it ;-) :

What about having "named slave servers", something as carte (maybe an
extension). A server knows it own name and knows how to connect to a
repository (or to a local set of jobs/transformations).

In one of the GUI's some more scheduling metadata would need to be
built in so you could assign a job to a named slave server and put a
schedule on it: every 15 minutes, every first of the month, ...

Upon starting a named server connects to the repository and downloads
the jobs that are assigned to it and starts executing (using Quartz
e.g. as scheduling component). You can monitor the jobs from a GUI.
Maybe some kind of local override would be required for variables, ...

The advantage I would see in this is to minimize OS scripting code
around Kettle. You could just build in a small script per project e.g.
in inetd (on Unix) or as a service (on Windows) that starts a named
slave server and it would start executing the "project" with mininum
non-kettle stuff.

It would require some changes in the repository, some scheduling magic,


You received this message because you are subscribed to the Google Groups "kettle-developers" group.
To post to this group, send email to kettle-developers (AT) googlegroups (DOT) com
To unsubscribe from this group, send email to kettle-developers-unsubscribe (AT) g...oups (DOT) com
For more options, visit this group at http://groups.google.com/group/kettle-developers?hl=en