Tag: batch job monitoring

Batch job monitoring overview

Batch job monitoring is a powerful part of SAP Focused Run system monitoring.

This blog will give you the overview of functions of batch job monitoring.

For setup of the batch job monitoring, you can read all details in this blog.

Questions that will be answered are:

  • How does SAP Focused Run job monitoring work?
  • What can I alert on?
  • Is maintenance needed for batch job monitoring?

Batch monitoring overview

For batch job monitoring start the Job Monitoring FIORI tile:

The batch job monitoring overview screen opens now:

You can now zoom into the job errors per system:

And you can zoom in per job now:

Go back to the previous screen to click on the alert generated:

During the setup of batch job monitoring (see this blog), you make the settings on alerting and who to inform when a batch job fails.

In the settings of batch job monitoring you can alert per job on:

  • Job cancellation
  • Job start delay
  • Job run time
  • Message codes in the job log

Maintenance

Maintenance on Job monitoring is required. In the initial setup the SM37 job schedule is read. But you still need to perform maintenance:

  • Jobs that no longer run need to be removed
  • Changed job behavior (for example run time length)
  • New jobs (these you need to add)
  • New insights into job criticality and more points to monitor (you typically start with cancelled jobs only, later you learn and fine tune per job)
  • Receivers of alerts change over time

For maintenance, you can open the batch job configuration screen and with the icons on the right see the alerts for changed or deleted jobs:

Reference

The full setup guide for batch job monitoring can be downloaded from the SAP Focused Run expert portal via this link.

Batch job monitoring setup

This blog will explain how to setup batch job monitoring in SAP Focused Run.

The usage of batch job monitoring is explained in the overview blog for batch job monitoring.

Questions that will be answered are:

  • How to setup batch job monitoring?
  • Can I monitor batch job start delay?
  • Can I monitor batch job duration?
  • Can I monitor if specific error messages occurred in the job log?
  • Can I alert on batch jobs?
  • Can I notify on batch jobs errors by sending mails?

Batch job monitoring setup

To setup batch job monitoring start the Batch job monitoring FIORI tile:

On the top right now open the configuration screen by clicking on the wheel icon:

The batch job groups now open. Select the pencil icon to change:

In the next screen press the New button to add a new Job group:

A batch job group can be set up in many ways:
1. You define one group for each system.
2. You define a group per functionality that crosses multiple systems.
3. Combination of the above. For example all technical jobs are clustered.
Think twice carefully on this setup before you start to implement. Changing between the concepts can be costing a lot of work.

Give the new group a name:

Give the new job group a good description and add the systems:

Save the definition.

First goto the third tab to make the default job settings:

The default job settings will be applied to any job added to the group. You can still fine tune per job later on. It is best to start with job status monitoring first. And only switch on the more advanced options later on for specific jobs only.

Save your settings again.

Now go to the jobs tab and press the add button to add new jobs:

In the selection screen that pops up now, it is very important to first select the system you want to read the jobs from, and then press go:

Focused run will now read the jobs from the remote system. In the list, select the jobs you want to monitor. You can select them mass as well. Then press ok.

Save your work.

The selected jobs are actively monitored now.

The jobs need to run first before they are visible in the batch job monitoring overview screen. Especially for the weekly and monthly running jobs.

Fine tuning the job monitoring: job errors

Fine tuning of the job monitoring can be done on the monitoring rules tab. First select the job to fine tune:

Now you can start to make deviations from the default.

In the example above the job duration monitoring was activated. Note here that you should set 2 thresholds: 1 for yellow and 1 for red.

The third option is job activity: this you switch on for jobs you need to have running at a certain frequently.

The fourth option is the job delay: some job really must be started in time without delay. Indicate here after which time an alert should be generated if the job does not start in time.

The last option is to check the job log messages. Here you can check for messages in the job log. Usage example: a job might have finished technically correct, so the job cancellation is not triggered. But inside the job log a functional error might occur. This error code you can pickup here and trigger an alert in the job monitoring.

Fine tuning the job monitoring: alerting

In the second block you can fine tune the alerting per job:

You can choose here to automatically confirm the alert if the system detects that the next batch job run went ok.

The other important fine tuning possible here is decide how many job failures in a sequence are needed before the alert is raised. For example, you have a job that runs every 10 minutes, but on average fails twice per day and is normally corrected the next run. In this case you can set the Relevant job instance counter to 2. This way 2 failed jobs in a row will alert. If it is recovered after a single wrong run there is no alert generated (here you must also fine tune the data collection frequency).

In the alert notification section you can assign an alert notification variant (this holds the mail template and receivers). In the default setup, you normally set the notification towards the basis team. But for some job(s) you might also want to notify a functional or business team.

Reference

The full setup guide for batch job monitoring can be downloaded from the SAP Focused Run expert portal via this link.