Lesson 3 of 6
Job and automation monitoring
9 min
Listen to this lesson
Batch jobs run the parts of the business nobody watches: overnight MRP, output processing, IDoc posting, billing, payment runs, period-end steps. When a job fails silently, the business finds out the next morning from a customer. Job & Automation Monitoring gives the run team one view of job executions across ABAP systems and cloud schedulers, with events when a job fails, ends with errors, runs too long or starts too late.
Sources and prerequisites
| Source | Examples | Prerequisite highlights |
|---|---|---|
| ABAP systems | SAP S/4HANA (Public and Private Edition), Business Suite 7, BW/4HANA, IBP | ST-PI 740 with SP16-20 and Note 3102288, or SP21+ and Note 3281776, or SP26+ and Note 3563104 |
| SAP BTP schedulers | SAP Job Scheduling service for Cloud Foundry or Kyma apps | Data push configured in the service |
| Automation services | SAP Build Process Automation, SAP Intelligent Robotic Process Automation | Data push configured in the service |
| Data services | SAP Datasphere tasks and task chains | Data push configured in the service |
Supported sources and prerequisites from SAP's setup page; check the current list for your release.
For ABAP jobs, the application log is often where the real result lives: a job can end green while the program wrote errors to SLG1. SAP's setup page describes linking application logs to job executions with transaction CRIT and the profile for auto-linking application log handles for jobs, with "Collect Job Monitoring" flagged. Do this; otherwise the monitor can only report technical success.
Two statuses, four event types
| Status or event | Meaning | Contoso use |
|---|---|---|
| Execution status | Technical outcome: red if aborted, yellow if finished with warnings in the job log | Every critical job |
| Application status | Business outcome from the application log: red for errors, yellow for warnings, grey without log | Billing, IDoc posting, payment run |
| Critical Runtime event | Run time above your threshold | MRP run for plant 1710 |
| Critical Delay event | Start delayed beyond your threshold | Output and IDoc jobs before the warehouse shift |
Statuses and events in Job & Automation Monitoring.
| Question | SM37 in each system | Job & Automation Monitoring |
|---|---|---|
| Scope | One ABAP system and client | ABAP systems plus cloud schedulers and automation services |
| Business result | Open the application log by hand | Application status rated from linked logs |
| Timing problems | Compare start and end times by eye | Critical runtime and delay events with thresholds |
| Who is told | Whoever looks | Alert, email, ticket or operation flow to a named owner |
| History | Job log retention per system | Hourly aggregates up to 750 days |
Good job monitoring depends on job hygiene. Contoso's rules: every critical job has a name with a Z_CMC_ prefix and its process (for example Z_CMC_BILLING_1710), runs under a technical batch user, writes an application log, and has a named business owner in the critical batch list. Jobs that change names every run, such as those generated by some applications, are grouped automatically; SAP's monitor lets you rename such groups so they stay recognizable.
Configuring events
- Open the configuration of Job & Automation Monitoring for the service (CMP) and confirm data collection is on.
- Create an event for critical execution status with a filter on job names, for example names starting with Z_CMC_BILLING.
- Create an event for critical application status on the same jobs; decide whether yellow should alert.
- Create critical runtime and delay events for the MRP and output jobs, using the baselines measured in CMR.
- Enable "Create Alert" and "Send Email To" with recipients from Notification Management; set the raise action to "at every occurrence" for jobs where each failure matters.
- Check housekeeping: single executions default to 2 days (1 to 7), hourly aggregates to 365 days (30 to 750), with at least the last 15 executions per job retained.
Two details catch teams out. Event filters combine values of the same parameter with OR and different parameters with AND, so a filter on job name "contains BILLING" and user "BATCH_SD" means both must match. And SAP notes that events are not raised for services in maintenance or disruption state in Business Service Management, which is useful during a release weekend and dangerous if someone forgets to end the maintenance window.
What would you do?
Every morning the billing job in CMP finishes green, but twice this month Maria found invoices missing for Great Lakes Fleet Services. The job's application log shows errors for those orders. What should Sam configure?
- Job & Automation Monitoring covers ABAP jobs and cloud schedulers and automation services.
- Execution status shows technical success; application status, via CRIT linking, shows business success.
- Use critical runtime and delay events with thresholds measured in pre-production.
- Events are suppressed for services in maintenance or disruption state; end maintenance windows on time.
Check your understanding · not graded
Lesson check
1. What does application status reflect for an ABAP job?
2. Which transaction links application logs to job executions?
3. Why might no alert fire for a failed job during a release?
4. Where should runtime thresholds come from?