Performing a direct upgrade of agents and domain managers

There are several methods you can choose from to upgrade your domain managers and agents.

The agent upgrade can be performed with minimal impact to scheduling activities. The agents are stopped for the shortest time necessary to perform the maintenance. Any active agent command-line interfaces and processes, such as conman, composer, netman, mailman, and batchman, to name a few, continue running. Any jobs already running when the upgrade process begins, continue to run as planned, however, no new jobs begin execution during this time. Once the upgrade is complete, the agent is restarted and quickly reconnects with its jobs. Any jobs that were actively running before the upgrade that have not yet completed, continue to run, and any jobs that successfully finished running during the upgrade procedure report a successful job status. An automatic backup and restore feature is in place in case of failure.

Because domain managers are agents, they are upgraded using the procedures described in this section.

If you choose to upgrade your environment top-down, then the agents get upgraded progressively after you have upgraded the master domain manager and its backup. This means that new features and enhancements are not available on all of your agents at the same time. If, instead, you choose to upgrade your environment bottom-up, then the agents are upgraded first, and new features and enhancements become available after the master domain manager and its backup have been upgraded.

Important: After upgrading your fault-tolerant agents, it might be necessary to manually update the security file on the fault-tolerant agents in your environment to add access to folders for all of the scheduling objects that can be defined or moved into folders. These updates are especially important if you plan to use the command line on the fault-tolerant agents to perform operations on the objects in folders. See Updating the security file for more information.
You can choose to upgrade your agents using any of the following methods:
twsinst script
A single line command that checks if processes or a command line is running before it starts. It saves disk space and RAM because it is not Java-based. See Upgrading agents and domain managers with twsinst and Upgrading agents on IBM i systems
Centralized agent update
Upgrade or update multiple fault-tolerant agent and dynamic agent instances at the same time. Download the fix pack installation package, or the eImage upgrade package to the master domain manager and then either run the installation on multiple agent instances or schedule the installation by creating and submitting a job to run. This upgrade method is not supported on z-centric agent instances. See Centralized agent update.

For a list of supported operating systems and requirements, see the System Requirements Document at HCL Workload Automation Detailed System Requirements.

When the upgrade procedure has completed successfully, the backup instance is deleted.

Note: The localopts file is not modified during the agent upgrade process. The file generated by the upgrade process is saved to the /config directory to maintain your custom values, if any. You can then merge the two files with your customized values and save the resulting file in the following path:
On Windows operating systems
<TWA_home>\TWS
On UNIX operating systems
<TWA_DATA_DIR>

When upgrading dynamic agents featuring both a local and a remote gateway, ensure you either upgrade the agent first and then the gateway or upgrade both at the same time.