Automated workload scheduling

For robust, reliable and secure batch job management in a networked environment, ROC Maestro for MPE is the solution. Easy installation and reliable functionality make this powerful solution the choice of thousands of MPE users. In addition to date, day of the week and time, ROC Maestro for MPE can schedule jobs anywhere on the network according to standard or customized business calendars. At the start of each processing day, ROC Maestro for MPE automatically selects all groups of jobs that run on that day, and carries forward uncompleted schedules from previous days. A schedule can use combinations of calendars to determine execution days, and multiple schedules can use the same calendars. Time constraints can be set for both jobs and schedules, specifying the time that execution should begin, or the time after which execution should not be attempted.

Job initiation can also be dependent on the successful completion of other jobs and schedules, on the existence of needed files on any node in the network, on user response to a ROC Maestro for MPE prompt, or on the availability of critical hardware or data resources.

The system can manage multiple schedules, each with its own set of constraints and dependencies. When there is a limit as to how many jobs can run simultaneously due to resource or administrator constraints. ROC Maestro for MPE selects which jobs to run based on an extensive priority scheme (effective range of 100).

The systems recovery facility ensures that aborted jobs are addressed automatically. After detecting failed jobs, ROC Maestro for MPE puts dependent jobs on hold while it initiates any combination of the following recovery actions: alert the operator, run a recovery job, stop processing, rerun the original job, or continue with the next job(s) in the queue. Maestro also easily incorporates and controls jobs it did not schedule to prevent overloading and conflicts.

ROC Maestro for MPE's sophisticated architecture provides multiple benefits. Fault tolerance guards against both system and network failures. If the central management system fails, a second system can take over. In case of a network failure, remote systems can continue launching jobs without instructions from the central management system. From a central system, ROC Maestro for MPE defines and monitors job scheduling, not only on MPE systems but also on any other networked UNIX or Windows NT system.

ROC Maestro for MPE allows system administrators to obtain up-to-the-minute information on the job status for each system in the network, with automated status message routing. Job statistics are logged automatically to a database, providing online access to a history of success, failure, minimum, maximum and average run times. A variety of reports, including histograms, calendar charts and cross-reference reports provide a clear picture of planned job activity and actual job results. A single console administers file and resource dependencies, system and schedule limits, and full inter-system job and schedule dependencies.

ROC Maestro for MPE's Signaler can send “alerts” when system errors or events occur. For example, a user can be notified if a critical job has failed or if output is ready. Signaler is designed to send remote users text messages via supported email, text pagers, cell phones, or websites.

Compatibility—Runs on all supported versions and many older versions of HP MPE/iX.