Monitoring and progress views

This is where you can find out why a job is behaving as it is, what job caused the situation, and what jobs are also affected.

M-Monitor Command

Use M-Monitor to see the upstream jobs of a job. An upstream job has to run before this job can, either because it's a triggering job or it's a predecessor. For example the upstream jobs for FUL3101 are FUL1103, FUL1205, FUL1102, FUL1101, and FUL0001H.

This list of upstream jobs have many of the same columns as the job instance displays. The panel below has been scrolled to the right to highlight the columns of particular interest here.

Job Instance Id

Along with the name this column has an * if this is the job the M-Monitor command was issued for, or a C if the job is a culprit job.

Ref Number, Trig./Repeat, Pre01 ...

These are numbers used to show the relationship between the jobs. Ref Number is a label for each job in the list. Trig./Repeat lists the reference number of the triggering job, or the iteration of the Repeating (Job) Group that this job is waiting on. The Pre01 (... Pre16) lists the reference numbers for the predecessor jobs.

Job Time

Job time is used to determine if a job is running longer than expected. PCS measures the time between when all the predecessor and triggering relationships of the job have been satisfied and the current time (or the job's completion time if it is already finished), plus a buffer factor. When this interval exceeds the job time, the job is considered a culprit.

In the above example, the Monitor command was issued for FUL3101. It is a culprit (C appended to the Job Instance Id column). It is triggered by FUL1103 (reference number 4 in the Trigger column), and has a predecessor of FUL1205 (reference number 5 in the Pre01 column).

FUL1103 and FUL1205 are both triggered by FUL1102, which is triggered by FUL1101, which is triggered by the header job FUL0001H.

You can, in turn, use the I-Info(UDF) and G-Progress commands on any of these jobs to get more information. Remember to refresh the screen to get up-to-date information.

G-Progress Command

The G-Progress command allows you to see your job in context of the rest of the workload. It answers the following questions:

• What are the preceding jobs?
• What are (just) the preceding jobs that are culprits?
• What are the affected, or downstream, jobs?
• Why does this job have an acceptable time of ...?
• Why is this job a culprit? OR not a culprit?
• Why is this job late? OR not late?
• Why hasn't this job completed?
  

Tab to the numbers in the All and To-Do columns for a list of preceding jobs, preceding culprit jobs or downstream affected jobs. The questions relevant to this jobs are also accessible from here.

Each of these views are explained below.

Preceding Jobs View

The Job Instance Id has * if the job is the focus of the progress command, and C if the job is a culprit.

The Reference Number is used in conjunction with the Trig./Repeat and Pre01 (...Pre16) to show the triggering and predecessor relationships amongst the jobs in the focused job's paths.

This view shows you the upstream jobs (both triggering and predecessors) and let's you drill down on any of them, using the G-Progress on them or S-Select to see the context for job in question.

The following panel gives the results for the S-Select Job Instance command for FUL3101. Notice the color-coding; this display highlights the immediate jobs - the jobs one relationship away from the job selected.

In this example FUL1103 is FUL3101's triggering job (in yellow), and FUL1205 is its predecessor.

Preceding Culprits Jobs View

This view shows you the upstream jobs (both triggering and predecessors) that are culprits and let's you drill down on any of them, using the G-Progress on them or look at their defaults and options and so on, similar to the Preceding Jobs View

Affected Jobs View

This view shows you the downstream jobs (both triggered and successors) and let's you drill down on any of them, using the G-Progress on them or S-Select to look at their immediate relationships, similar to the Preceding Jobs View.

The affected jobs for FUL3101 are its downstream jobs: FUL3104T is triggered by FUL3103 (reference number 3), and both FUL3103 and FUL3102T are triggered by FUL3101 (reference number 1).

The S - Select Job Instance line command gives insight into the related jobs. Here's information for FUL3103:

Here FUL3101 is the triggering job for FUL3103 and FUL3104T is triggered by FUL3103.

Acceptable Time Explanation

One of the questions on the Progress panel is "Why this Job Instance has an Acceptable time of ...?" This view shows how the acceptable time was calculated and what factors affected this calculation.

In this example the acceptable time for FUL3104T is inherited from the application's acceptable time. The acceptable time for FUL3103 triggers FUL3104T and must start two minutes earlier to allow the trailer job to complete on time. In the same way, FUL3101 is upstream from FUL3103 and it must complete three minutes earlier to give FUL3103 time to complete on time.

Battle Plan Ending Time (i.e., the end of the production cycle), Application Acceptable Time, Job Tree Acceptable Time, and User-defined Value (i.e. the acceptable time specified at the job level, i.e., L4 or L8) are all specified in the defaults and options sub-dialog and inherited according to the PCS precedence order.

However Triggered Job Instance and Successor Job Instance both tell you that a downstream job has an acceptable time that can't be met unless this job completes earlier than it would otherwise. In this case the acceptable time calculation is based on the propagation process from the trailing jobs backward through the trees to make sure all jobs can complete on time.

While these are the most common reasons, there are others covered in the elective feature they are associated with. If they apply to your implementation, see Shifting

Definition of Lateness Temporarily, Repeating Group Options, Acceptable Time Considerations, and Crossing Production Cycle Boundaries.

Culprit Explanation

Another question on the Progress panel is "Why this Job Instance is a Culprit ?" This view shows culprit time was calculated and what factors affected this calculation.

Recall Job time is used to determine if a job is running longer than expected. PCS measures the time between when all the predecessor and triggering relationships of the job have been satisfied and the current time (or the job's completion time if it is already finished), plus a buffer factor. When this interval exceeds the job time from experience data, the job is considered a culprit.

The culprit time for FUL3101 is the completion time of its last dependent job plus the average job time plus a buffer calculated over 37 samples. FUL3101 became a culprit at 10:21 and has been for over five hours.

Lateness Explanation

Another question on the Progress panel is "Why this Job Instance is late?" The response explains what PCS knows about how this job's ending time was determined (by definition it's later than its acceptable time), whereas the acceptable time explanation focusses on which settings determined the final calculated acceptable value.

The main causes for a job to be late are an unrealistic acceptable time, being subject to an upstream culprit job, and running longer than expected (and perhaps becoming a culprit itself).

What's holding up the job

The last question on the Progress panel is "Why this Job Instance has not completed?" This view summarizes the state of the job and gives options to determine more information, including some predetermined CA 7 commands.

Using 6 UDF Info for the Job Instance results in the following information.

Since the job is held and JB is in red, tabbing to JB for more information shows the job is held for DOCT.FUL3101 binding agent (highlighted).

Indeed once the binding agent is activated FUL3101 executes as does the rest of the jobs in its paths.