Some great transformers built right into FME Server can help you maximize the productivity of your batch processing. By making it easy to automate running a set of transformation jobs in sequence, these transformers give you more flexibility to prepare your data exactly how it’s needed.
Using FME Server you can run a “control” workspace that queues up other transformation workflows to be performed on the server, either concurrently or in a specific sequence.
When the jobs complete, you can
have FME automatically push the output data wherever you want it – either as its own output (like normal) or as input data for another transformation job.
So instead of manually running a group of transformation jobs one after the other, FME Server enables you to queue up multiple transformation workflows to run in logical stages without human intervention.
Let’s look at a couple of examples.
The simplest involves a control workspace that queues up multiple transformation jobs to run. These jobs can be set to run in parallel on any available FME engines, or they can be set to wait for a previous transformation to be complete before they begin. This allows a step-by-step approach to data transformation without requiring a person sitting behind a screen to commence each stage.
How would you do this? Well it’s actually quite simple. Your control workspace uses an FMEServerJobSubmitter transformer to queue up the transformation jobs to run. These transformation jobs are simply other workspaces.
To tell the next set of transformation jobs to wait for the previous jobs to complete successfully, you use the FMEServerJobWaiter transformer. Place it in your control workspace and the second set of jobs will wait patiently for the results of the first set to be in before they run. Use it again to have a third set of jobs wait for the second set to complete successfully, and so on.
Now let’s look at another example.
Instead of just queuing up the next set of transformation jobs to run, you can use the output of previous transformation jobs as input for the next set of transformation jobs. To do this, you simply use a transformer such as the FeatureReader to bring the output back into the control workspace.
(Never heard of the FeatureReader transformer? Learn more about it in the article starting on page 2.)
In both of these examples, you gain the increased productivity of the next generation of batch processing. Plus, you gain the advantages of FME Server’s enterprise-ready features to ensure that your transformation project goes smoothly: job scaling over multiple engines and machines, job queuing, job reporting, fault tolerance, failover protection, and more.
Learn more tips and tricks for using FME Server in the tutorial.