HCL Workload Automation, Version 9.4

Plan management basic concepts

The production plan contains information about the jobs to run, on which fault-tolerant agent, and what dependencies must be satisfied before each job is launched. HCL Workload Automation creates the production plan starting from the modeling data stored in the database and from an intermediate plan called the preproduction plan. The preproduction plan is automatically created and managed by the product. To avoid problems, the database is locked during the generation of the plan, and is unlocked when the generation completes or if an error condition occurs. The preproduction plan is used to identify in advance the job stream instances and the external follows job stream dependencies involved in a specified time-window.

You use the JnextPlan script on the master domain manager to generate the production plan and distribute it across the HCL Workload Automation network.

You can run the JnextPlan command from a command prompt shell on the master domain manager if you are one of the following users:
  • The TWS_user user for which you installed the product on that machine, if not disabled by the settings that are defined in the security file.
  • Root on UNIX operating systems or Administrator on Windows operating systems, if not disabled by the settings that are defined in the security file.

For additional information on the JnextPlan script, refer to Creating and extending the production plan.

To generate and start a new production plan HCL Workload Automation performs the following steps:
  1. Updates the preproduction plan with the objects defined in the database that were added or updated since the last time the plan was created or extended.
  2. Retrieves from the preproduction plan the information about the job streams to run in the specified time period and saves it in an intermediate production plan.
  3. Includes in the new production plan the uncompleted job streams from the previous production plan.
  4. Creates the new production plan and stores it in a file named Symphony. The plan data is also replicated in the database.
  5. Distributes a copy of the Symphony file to the workstations involved in the new product plan processing.
  6. Logs all the statistics of the previous production plan into an archive
  7. Updates the job stream states in the preproduction plan.

The copy of the newly generated Symphony file is deployed starting from the top domain's fault-tolerant agents and domain managers of the child domains and down the tree to all subordinate domains.

Each fault-tolerant agent and domain manager that receives the new Symphony file, archives the previous Symphony to Symphony.last in the path <TWA_home>/TWS/, so that a backup copy is maintained. This permits viewing of the previous Symphony data in case there were any message updates on the job and job stream states that were lost between the agent and its master domain manager.

Each fault-tolerant agent that receives the production plan can continue processing even if the network connection to its domain manager goes down.

At each destination fault-tolerant agent the HCL Workload Automation processes perform the following actions to manage job processing:
  1. Access the copy of the Symphony file and read the instructions about which jobs to run.
  2. Make calls to the operating system to launch jobs as required.
  3. Update its copy of the Symphony file with the job processing results and send notification back to the master domain manager and to all full status fault-tolerant agents. The original copy of the Symphony file stored on the master domain manager and the copies stored on the backup master domain managers, if defined, are updated accordingly.

This means that during job processing, each fault-tolerant agent has its own copy of the Symphony file updated with the information about the jobs it is running (or that are running in its domain and child domains if the fault-tolerant agent is full-status or a domain manager). Also the master domain manager (and backup master domain manager if defined) has the copy of the Symphony file that contains all updates coming from all fault-tolerant agents. In this way the Symphony file on the master domain manager is kept up-to-date with the jobs that must be run, those that are running, and those that have completed.

The processing that occurs on each workstation involved in the current production plan activities is described in more detail in HCL Workload Automation workstation processes.

Note: While the current production plan is in process, any changes you make to the plan using conman do not affect the definitions in the database, but the replica of the plan data in the database is updated with the changes. Subsequent updates to job instances in the plan are supported, but do not affect the job definitions in the database. Changes to the objects in the database do not affect the plan until the production plan is extended or created again using the JnextPlan script or planman command-line interface. Updates to objects in the database do not affect instances of those objects already in the production plan.