Lesson 3 of 5
Operations capabilities
7 min
Listen to this lesson
Go-live is the start of a different job. Once Contoso's S/4HANA system is productive, someone has to know quickly when an interface fails, a nightly job runs late, or order-to-cash slows down. Cloud ALM for Operations is SAP's answer: a set of monitoring and alerting apps that cover cloud services and systems across the landscape.
The operations apps
| Capability | What it watches | Contoso example |
|---|---|---|
| Health Monitoring | Technical health of cloud services and systems, with health checks and trends | Availability and resource trends of the production S/4HANA system |
| Integration & Exception Monitoring | Interfaces and messages end to end, plus application exceptions | A failed EDI order from Great Lakes Fleet Services |
| Job & Automation Monitoring | Background jobs and automations: status, runtime and start delays | The nightly MRP run for plant 1710 finishes late |
| Business Process Monitoring | Business process health and anomalies, using process KPIs | Sales orders stuck without deliveries in sales area 1710 / 10 / 00 |
| Real User Monitoring and Synthetic User Monitoring | Response times experienced by real users, and scripted availability checks | Order entry is slow for Chicago users |
| Configuration & Security Analysis | Configuration changes over time and settings against security recommendations | A security parameter changed in production |
| Business Service Management | Technical services grouped as business services, with availability against agreed levels | "Order to cash" service availability for the month |
| Intelligent Event Processing | Events from different sources, correlated and routed to the right handler | One alert instead of twenty for the same outage |
Cloud ALM for Operations capabilities as SAP lists them. Names may evolve with SAP's continuous releases.
The operations apps are not switched on automatically. Systems and services are registered in Landscape Management (cloud services on your customer number appear automatically; on-premise systems are added), and then data collection is activated per use case. A registered system shows nothing in Job & Automation Monitoring until someone enables it.
- RegisterCloud services appear in Landscape Management automatically; on-premise systems such as Contoso's S/4HANA are added
- ActivateData collection is switched on per use case, for example Job & Automation Monitoring for the production system
- ConfigureThresholds, alerts and recipients are set for each monitored object
- MonitorHealth, interfaces, jobs and processes show data; alerts reach the right team
Why testers should care about operations
Operations monitoring looks like a Basis and support topic, but it matters to quality in three ways. First, many defects found in hypercare are integration or job failures, and these apps are where they first appear. Second, the same monitors can be set up in the quality system before go-live, so integration failures during system integration testing are visible to the test team, not just the interface developer. Third, the operations data after go-live is the best evidence that the processes you signed off are actually running.
Implementation and operations in one tool
Because both areas sit in one tenant, a defect found in hypercare can be traced back to the process, requirement and test case it relates to. That closes the loop between "what we tested" and "what broke in production", and it tells the test lead which regression tests to strengthen before the next release.
A practical habit for Contoso: when a hypercare incident turns out to be a gap in testing, add a test case for it in Test Preparation, link it to the same solution process, and put it in the next regression plan. Over a few releases the regression pack grows from real production evidence rather than guesswork, and each fix leaves a test behind it.
- Hypercare incidentMonitoring or a key user reports a failure in production
- Trace backFind the process, requirement and test case it relates to
- Testing gap?If no test covered the case, add a test case in Test Preparation
- Link and planLink it to the same solution process and add it to the next regression plan
- Stronger regressionEach fix leaves a test behind for the next release
What would you do?
During Contoso's SIT cycle 2, several outbound invoice messages to Great Lakes Fleet Services fail silently. Testers only notice days later. What would most improve this in the next cycle?
- Cloud ALM for Operations covers health, integration and exceptions, jobs and automations, business processes, user experience, configuration and security, business services and event processing.
- Systems are registered in Landscape Management and data collection is activated per use case.
- Monitoring the quality system during testing makes integration and job failures visible to testers.
- One tenant for implementation and operations links production issues back to the tested processes.
Check your understanding · not graded
Lesson check
1. Which operations capability watches interfaces and messages end to end?
2. A nightly MRP job keeps finishing late. Which capability is designed to alert on that?
3. What must happen before a registered system shows monitoring data?
4. Why might a test lead want monitoring in the quality system?