Please click on the links below for elaborations of each of the features:
The Phantom Job Processor (PJP) was designed to be run with Columbia Ultimate’s (CU) “The Collector System” (TCS). Anyone familiar with CU’s user interface conventions will be able to use the PJP screens with no problem.
The PJP system is marketed by CU as an approved add-on for TCS.
Both the jBase and UniVerse software platforms contain the functionality to run programs in the background. The PJP makes use of this inherent functionality and gives it a user-friendly control console.
When jobs are being executed by PJP, they are using neither a CU / TCS port license nor a jBase / UniVerse port license.
PJP supports multiple logons. Within each logon, there is a one-to-many relationship which allows multiple phantoms to co-exist. Within each phantom, there is another one-to-many relationship with the capacity to schedule multiple processes.
Many agencies have to schedule operations staff to come in at times either before or after the normal business day to run large jobs. These are typically jobs which can’t be run while agents are active. PJP launches such jobs according to a pre-defined schedule and eliminates the need for having people work very early or very late in the day.
The screens of the Phantom Console program were designed using the same user input conventions of the TCS system. Anyone familiar with how to use TCS screens will be able to easily use the PJP interface.
Processes can be run according to very flexible scheduling parameters.
Additionally, there is the capability to have the PJP software automatically shut down for planned system outages.
A set of screens are built into the Phantom Console to display scheduled jobs in a flexible format so that the PJP admin can evaluate his whole set (or a specific sub-set) of scheduled processes. A front-end parameter entry screen allows the admin to choose which processes to display and the order in which to display them.
There are 2 monitoring programs which are run by the operating system (either cron for unix or scheduled tasks for Windows) which are set up to run hourly.
One of these programs (PHANTOM.MONITOR) looks for problem phantoms which haven’t “checked in” in the expected time frame. If potential problems are detected, an email is sent to the PJP administrator, alerting them of the situation.
The other program (PHANTOM.RESUME) runs one time for each logon, looking for phantoms which were shut down according to plan and for which it is time to re-launch them. If that situation is detected, the phantoms are launched automatically.
The Phantom Console provides the phantom administrator with access to an event log for each phantom. The log shows the current status plus a recap of the historical events including phantom events (start, stop, hung, killed), the dates and times that each process was launched and the elapsed time that it took for the process to run. This is a valuable tool for analysis leading to modifying schedules and also for troubleshooting.
PJP interfaces with email services on your host computer (sendmail or mail for unix; blat for Windows) to send emails to single or multiple recipients to notify them about process completions and/or phantom events (start, stop, hung, or killed). The list of email recipients is configurable by the phantom administrator using the Phantom Console program.
PJP is currently supported on the following server platforms:
Starting with PJP release 1.05, there is the capability of defining pre-built Easy Out reports as phantom processes. Output from an Easy Out report may be routed either to the printer or to a sequential file written to a folder on your TCS host.
Copyright 2011, Cubs Consulting, Inc.