DataStage Job Run-time Architecture
During the start phase of a parallel job several processes and verification's are performed. Here is a list of the most relevant: 1. The Parallel Engine starts the Conductor process and some other processes including job monitor process. 2. The Parallel Engine builds the score (the execution plan) of the job. 3. The Parallel Engine verifies the input and output schema of every stage. The larger the schema the longer this step takes. 4. The Parallel Engine verifies that the settings for every stage (operator) are valid. If a job has stages that interact with databases, DataStage connects to each database and if needed verifies that the database is configured properly to work with parallel processes. 5. The Parallel Engine creates a copy of the job design into disc. 6. The Parallel Engine connects to (remote) servers and starts Section Leader processes. Communication channels are created between Section Leader processes and the Conductor Process. 7. Section Leader...
