Add a Job Interactively

In this tutorial, we will build a simple flow graph one job at a time, interactively from the command line. This is to demonstrate the basic building blocks for adding jobs to a flow graph, and how we can monitor a flow using the FlowTracer GUI program "vovconsole". This is not to demonstrate production techniques for building a flow.

You will register jobs interactively with FlowTracer to have FlowTracer build its flow graph one job at a time. This will make it easy to see what is going on. In a production situation, jobs are not registered into the flow graph interactively. Instead, the jobs are registered into the flow graph by way of batch scripts or by processing control files. The batch style of registering jobs will be shown later in the tutorial.

Objective

  • Verify that FlowTracer is properly setup.
  • Become familiar with named environments.
  • Learn more about the concept of runtime tracing.
  • Become comfortable using the FlowTracer GUI for managing the jobs.

Use the "cp" Program to Emulate a Tool

You will use the UNIX copy program cp to emulate a more useful tool having an input and output. The cp command comes with UNIX, and FlowTracer supplies a script file called cp.bat which allows you to also run this tutorial on Windows.

The UNIX program cp reads an input file and copies it to an output file. It can be used for a very simple job to create a backup of a file.
cp input-file output-file
Consider a job that is a typical computer task using a tool named TRANSFORM:
Job 1: TRANSFORM source-file expanded-file

This job is a generic one that reflects what most tools do. It reads an input file and generates an output file based on it.

We will use the cp program during this tutorial as a fast and low cost tool to demonstrate how to register jobs into the flow graph, and how to monitor and control the work of FlowTracer, the flow manager.

Here is the above TRANSFORM job emulated using the UNIX cp command.
Job 1: cp input-file output-file

The goal of the job is to create the output file. The output file depends on the tool to generate it from the input file. If the input file changes, then the output file is out of date, and is INVALID using flow terminology. When the input file is changed, then the job's goal to create the output file is triggered. To reach the goal, the tool must run, process the input, and create a version of the output that is up to date. This makes the output file VALID.