3.14.3 Job monitoring function

Organization of this subsection

(1) Overview

In conjunction with JP1/AJS3 - Manager 13-50 and later, by monitoring JP1/AJS3 job information as metric, you can visualize the transition of the execution time of the root jobnet by predictive detection and anomaly detection of performance issues of the root jobnet execution time and visualization by integrated operation viewer. Trend data using JP1/AJS3 root jobnet execution time as a metric can be stored in the trend data management DB of JP1/IM - Manager and displayed and monitored on the Trends and Dashboards tabs of the integrated operations viewer.

Metric of JP1/AJS3 job-information displayed in integrated operation viewer is defined in metric definition file (metrics_ajs_rootjobnet.conf) of JP1/AJS3.

To Page Top

(2) Prerequisites

The following prerequisites apply when using the job monitoring function:

The JP1/AJS3 - Manager version to be linked must be 13-50 or later.
Trend data using the root jobnet execution time linked with JP1/AJS3 as a metric must be stored in the Trend data Management Database of the Intelligent Integrated Management Database in JP1/IM - Manager.^#
Trend data must be able to be displayed on the Dashboard tab or Trend tab of the integrated operation viewer.^#

#: JP1/AJS3 linkage settings must be performed and JP1/AJS3 root jobnet execution must be completed.

To Page Top

(3) Available version combinations (JP1/IM - Manager and JP1/IM - Agent connectivity)

The following shows a combination of versions that can use the job monitoring function.

JP1/IM - Manager 13-50 or later
JP1/IM - Agent 13-50 or later

To Page Top

(4) Use Case

(a) Monitor the execution time of JP1/AJS3 root jobnet with prediction detection alerts to proactively respond to increased execution time

In the operation of the job execution infrastructure using JP1/AJS3, the automated batch operation is executed as a root jobnet. It is important that batch operations be completed on time. However, if the batch operation is performed for a long time, the execution time may gradually increase and the processing of the operation may not be completed by the fixed time. Therefore, the execution time is monitored to ensure that it is within the expected range.

By monitoring the delay in the execution time of JP1/AJS3 root jobnet with prediction detection alerts and visualizing the transition of the root jobnet execution time with integrated operation viewer, you can cope with the increase in the execution time before the business is not completed by the fixed time.

For details about monitoring the duration of a JP1/AJS3 root jobnet, see Setup for linking JP1/IM3 in the JP1/Automatic Job Management System 3 Linkage Guide.

(b) Show a graph of the dashboard trend panel for metric with the range vector selector specified in promql of metric definition file

For details on specifying the range vector selector "$stepTime{minSeconds="minimum-seconds"}", see Consolidation display of trend data with dynamic range vectors in 3.15.6(4)(c) About Performance Data to Retrieve.

By specifying "$stepTime{minSeconds="minimum-seconds"}" as the range vector selector in promql of metric definition file, you can dynamically calculate and display trend data for the range vector to include all data according to the displayed range of trend data in the dashboard.

For example, in the trend panel of a dashboard, suppose you want to change the display range of the graph to a longer range than the default. In this case, instead of specifying a fixed value (1d) as the range vector selector, you can specify "$stepTime{minSeconds=\"86400\"}" to graphically display the correct value. Specifically, if you look at the graph in the dashboard trends panel for the average root jobnet execution time trend for 49 days every seven days (one week), the values displayed in the graph are as follows:

- Conditions for displaying graphs

Metrics to display: Average value of JP1/AJS3 "Root jobnet execution time"
"Root jobnet execution time" load: High load every Saturday and Sunday for 48 hours
Graph display range: 2025/01/08 (Wednesday) 00:00:00 to 2025/02/19 (Wednesday) 00:00:00 (42 days) ^#

#: 49 days of data from 2025/01/01 00:00:00 to 2025/02/19 00:00:00 are displayed.
Number of plots in graph: 7
Graph display interval: 7 days
Trend data storage interval: 1 day

- Promql settings in the metric definition file and values displayed in the graph

When a fixed value (1d) is specified as the range vector selector
- Promql of metric definition file
  
  (avg_over_time(ajs_rootjobnet_exectime[1d]) and $jp1im_TrendData_labels)
- Values displayed in the graph
  
  The plotted data is calculated based on the data for the last day of Wednesday 00:00:00 (the "root jobnet execution time" value run every Tuesday), so the graph is displayed with the calculated value without using the high load value.
When "$stepTime{minSeconds=\"86400\"}" is specified as the range vector selector
- Promql of metric definition file
  
  (avg_over_time(ajs_rootjobnet_exectime[$stepTime{minSeconds=\"86400\"}]) and $jp1im_TrendData_labels)
- Values displayed in the graph
  
  The plotted data is calculated based on the data for the week immediately before 00:00:00 every Wednesday, so the graph is displayed with the value calculated using the value with a high load. You can display the graph with more accurate values than if the range vector selector is a fixed value.

To Page Top