Enterprise Applications
Development Guidelines and Procedures
Job Processing #1
Job Setup
- Job Names
- Work Requests
- Analyst Jobs
- Priority Standards
- Input Priorities (Batch Jobs)
- Output Priorities (Spool files)
The session name in all stream jobs must be the name of the stream job file itself. This convention allows one to quickly find the creating stream job for any executing job. Decisions can then be more easily reached concerning whether an executing job must be aborted for backup or other system problems (See Appendix E).
Jobs submitted through MPEX should be named using the initials of the analyst submitting the job. See the MPEX manual for instructions on changing the default (See the %SUBMIT command).
All jobs requiring computer operator intervention to process such as tape jobs,
special forms, etc., should be initiated via a work request. The customer should send
the work request to Computer Operations via email indicating the job number and any special
processing required. A web form has been developed to simplify the submission of a
Work Request.
On occasion it will be a temptation for an analyst to perform production processing
rather than turn it over to the department. This is strictly prohibited. Departments and
customers must be required to accept this responsibility.
Input and output priorities should be used to help speed production processing and
avoid confusion. The following standards are to be utilized:
| HIPRI: | Special Use. For System Manager and System Operator emergency use only. |
| 0 | Special Use. Previous job on a workorder has aborted. Do not run this job without analyst approval. |
| 1 | After Hours. Low priority for overnight processing. |
| 2 | Work Request. Check work request for special instructions before processing, utilized for tape jobs, etc.. This priority will also be used by any customer submitting their own work request. If no work request is found or received the job will be deferred with an input priority of zero and will not be processed. |
| 3 | Earliest Convenience. Short running jobs which may be run during the day at the convenience of Computer Operations. |
| 4 | After Backup Has Started. Run after the online backup has been initiated and the system has been made available again, but before 7:00 a.m. For any job not started by 7:00 a.m., the operator will contact the analyst of the system to determine whether the job should be kept for the next night's processing, run during the day, or purged. If the analyst is unavailable, the job will be flushed from the queue and a speed letter written to the analyst to indicate the job was not run due to lack of time. |
| 5 | To be used for jobs that need exclusive access to a specific database which is also accessed by PowerHouse Web -- usually, these will be short-running Adager jobs. These jobs are to be run after the Backup has processed its synchronization point, but before the PHWEBJ,MANAGER.COGNOS job is allowed back on. They can run concurrently, as each job is associated with a discrete system, except that no jobs with identical system codes should run together (e.g. no two 'SI' jobs with an inpri=5 should run together, but an 'SI' and an 'AI' job can run together.) |
| 6 | Not currently utilized. |
| 7 | Special Use.
Interdependent jobs which do not require a work request but must be run as soon as possible.
Should one of the jobs bomb all remaining jobs should not be launched without customer approval.
Each series of jobs must be preceeded by a job that prints a tracking sheet listing the jobs
to run and their sequence.
Any jobs with a priority of 7 not having a corresponding tracking sheet will not be processed. (Note: Most often used with payroll processing) |
| 8 | Earliest Opportunity. No computer operator intervention is required [default]. Jobs will run in the order of submission. No two jobs with the same system code or from the same logon USER.ACCOUNT should run together. |
| 9 | Emergency Use. To be utilized for emergency problem resolution, with discretion, and in communication with Computer Operations. |
| 10+ | Do Not Use! Will conflict with current Computer Operations procedures. |
| 0 | Special Use. Do not print or delete spool file. Output is from an aborted job. |
| 1 | Special forms. Will not be printed without a work request. Use when bursting, trimming, or other output instructions are required. Output with this priority must be placed with the work request and logged for pickup by the appropriate department. |
| 2 | Special Handling. For multiple copies or large reports. If not used for multiple copy reports the output may be purged. |
| 3 | Special Tracking. Output created by a job ran from a work request. Must be placed with the work request before pickup by the appropriate department. However, unlike output priority 1, no forms or special handling are required. Output created by jobs on a work request not deferred with a priority of 1 or 3 will not be kept with the work request. |
| 4 | Do not print. For review only. |
| 8 | Earliest Opportunity. No computer operator intervention required (default). |
| 9 | Permission Only. To be used for emergency problem resolution, with discretion, and in contact with Computer Operations. |
| 10 | Same as 9. |
| 11+ | Do Not Use! Will conflict with current Computer Operations procedures. |
- Jobs to be run by the end user.
- Files must be moved into the appropriate production groups. Batch jobs should be placed in the JOBS.hpaccount group unless the customer requires the ability to modify the job, in which case the job should be placed in the customer's HP logon group. On-line Powerhouse programs should be placed into the PRODQ.hpaccount group unless it is a QUIZ report which the customer must modify. QTP programs must not be placed in the customer's group, but should be placed either in JOBS or PRODQ depending on the program type (batch vs on-line). Compiled COBOL programs and Powerhouse dictionaries should be placed in the PROD.hpaccount group (For more information see Section 8, Part A, Number 4).
- A customer documentation file must be established and placed in the DOC.hpaccount group for all on-line processes not adequately internally documented for the customer.
- Descriptions of all new files other than jobs placed in the Administrative Data Processing Master Schedule will have descriptions placed in Fileindex. Descriptions of jobs will be automatically transferred from Master Schedule to Fileindex. A description from any flat file will be automatically transferred to Fileindex if a short one-line description is prefaced with '~D' and a space. The '~D' feature will be used in all program source files (See documentation examples: Section 8, Part B).
- Documentation must be prepared and distributed as per instructions in Section 8, Part B.
- The user must be instructed how to properly use the process and to display or print the available user documentation.
- The job should be scheduled by the user, if appropriate.
Any job, on-line or batch, ready to be placed into production must be prepared for production using the following procedures:
Customers may schedule jobs through the Administrative Data Processing Master Schedule system. Analyst assistance should be provided to assist customers who need to maintain schedules for many jobs and wish to utilize this system.