c D SCHEDULE Automated Job Submission System Guide and Reference Manual D SCHEDULE
Automated Job Submission System
Guide and Reference Manual
J 1.3 Main features
Some of the main features are:
I - A full MOTIF interface with live monitor displays and pull down menus.C
- A command line interface are provided with j complete help for all possible commands and qualifiers. See Chapter 4 for more on the line commands.I
- A full Menu/Forms interface 2
for 0 ANSI VT terminals and DECWindows workstations.I - Support for all wildcards. They can be used when I describing job names and other job elements. This allows you to modify & multiple jobs with a single command.E
- A MOTIF Graphic job editor can be used with F workstations. The main component of the Editor/Monitor, this editor G lets you define and interconnect jobs using a line/box display and a ) mouse to create the needed connections. D
- Interval and named day scheduling. You can J specify a job start time and schedule it to run at regular intervals by F using the name of the day or specifying at what interval to run the T job. See Section 2.1.1 for more on scheduling.G
- Job scheduling can be calendar based. You can I create calendars that recognize holidays and weekends when jobs are to H run, and other calendars that mark specific days on which jobs are to F run. You then specify which calendar each job is to follow. You can E also create fiscal calendars for jobs concerning financial matters r (such as quarterly financial reports). See Section 2.1.1.3 for more on calendar based scheduling.?
- Two types of interdependencies can be E established. The forward going connection is called an "initiate". E Initiates are triggered when a job completes. Associated with each F initiate can be a condition or severity level at which the next job J should be triggered. In this way, different branches can be taken based n on the success/failure of a particular job. See Section 2.1.8 for more B on initiates. The second connection is called a "prerequisite." I Prerequisites hold up a job, preventing it from being submitted into a k VMS execution batch queue. See Section 2.1.10 for more on prerequisites.E
- Resource dependencies can be specified that I cause jobs to wait until a specified file and/or a specified amount of ) disk space is present before executing. G
- Jobs can be grouped into "Job Sets". Multiple G copies of the same job stream can be initiated to run simultaneously H with different data sets, with complete independence between each set.I
- SCHEDULE uses the standard VMS E batch queue facilities to execute jobs. There is no need to change A development command files for production use within the system. H
- The standard eight parameters can be passed to D any job. Along with these, there are two more parameters that are & maintained across the whole job set.<
- Directory organization - The entire J SCHEDULE database is organized into directories. Any number of I directories can be created. Access to each directory is controlled and o any number of concurrent users can use the system. See Section 2.3.3 for * more on directories in SCHEDULEB
- Access controlled - All access to the job F database is controlled via the standard UIC based access or through VMS standard ACL's . C
- Support for any combination of CPU's in a I cluster is supported. For each job, you can specify the queue and user F name under which the job is to run so that the job execution can be G controlled using the standard VMS batch queueing structure. F You can use generic queues to distribute jobs over many nodes. If a E node fails, the job can automatically be restarted on another node serviced by the generic queue.E
- Networks supported - Support is provided for G systems that are networked together but not clustered. You can run a J job on another node from the local node, tell a local job to wait for a G remote job before starting, and have local jobs initiate remote jobs. J
- Distributed or centralized database support - The C SCHEDULE database can be configured across a network of H systems either as a centralized network with satellites or as a group of peer-to-peer databases.G
- Automatic restarts - You can specify that jobs H are to be automatically restarted upon failure, and define the number I of job restarts to occur before you are notified of a job failure. You E can also design jobs to have restart points so that, when a job is 2 restarted, it restarts at the appropriate point.C
- Notification messages are sent at several J points. For each job, you can indicate which user(s) are to be notified E when a job reaches a specified state (submitted, started, delayed, C completed or failed). Notification can be by E-mail or broadcast C messages. You can define your own messages to be included in the notification. C
- Distribution lists - For each job, you can H specify a list of standard PRINT (other commands such as COPY or MAIL J will also work) commands to distribute the various types of print files J produced by a job. This list is independent of the actual job commands, G and it can readily be changed without affecting the logic of the job itself. B
- Jobs can be monitored. SCHEDULE H provides several ways to monitor jobs in progress over the cluster or C network. From the command line, you can run a continuous display J monitor showing job progression, or enter DCL commands for snapshots of H the scheduling queues and their entries. For VT and workstations, you H can select the Job Monitor Window which will maintain a window on the H screen that tracks all activity. Or for a workstation, you can invoke E the Editor/Monitor for graphic display for the job network and its current state.G
- Report writer - A large collection of standard J reports are included. You can select from the existing group of reports @ or create your own by customizing the report definitions. See O Chapter 14 for more on the report writer.
/ N Note
G 2 The Menu/Forms interface is N implemented using the MCL Subsystem also available from your distributor. |