Will a work plan or job schedule be sent to a client, and can this plan be processed without a JobScheduler server being available?
JobScheduler itself does not use a predefined work plan or job schedule - it is not necessary. Similarly, we do not differentiate between server and client JobSchedulers.
- Every JobScheduler works with its own configuration (which it can receive locally or remotely from another JobScheduler Supervisor) and is thereby independent of other JobSchedulers at run-time.
- At the same time, each JobScheduler is able to control processes on other JobScheduler instances (remote execution) or on other servers (Agentless Scheduling). In this manner, each JobScheduler instance can be seen as operating as a server and/or as a client.
JobScheduler Agents, however, are exceptions: they do not have their own configuration and their jobs are controlled remotely by another JobScheduler in a Master / Agent Cluster. This arrangement comes closest to the classical client/server architecture.
The robustness of JobScheduler against failure can be ensured by the installation of a Passive Cluster or Active Cluster.