Lesson 2 of 5
Monitoring and triage
8 min
Hypercare has two sources of information: what users report, and what the system shows. Good hypercare uses both. Monitoring finds failing interfaces and jobs before users notice; triage sorts every reported issue quickly so it reaches the right person with the right priority.
Monitor the system, not just the tickets
SAP Cloud ALM for Operations provides monitoring applications that are useful from the first hour: Business Process Monitoring, Integration and Exception Monitoring, Job and Automation Monitoring, Health Monitoring, and user and performance monitoring such as Real User Monitoring. Set them up and test the alerts during Deploy, so that on go-live day failed IDocs, stuck interface messages and failed batch jobs raise alerts to named people.
The top layer is the most important and the most often forgotten. For the first weeks, Contoso compares each day's count of sales orders, deliveries, billing documents and goods receipts with a typical day in ECC. When deliveries for Great Lakes Fleet Services rose but billing documents did not, the gap revealed a billing block before anyone raised a ticket.
Triage categories
Triage is the most valuable skill in hypercare: quickly sorting each issue into the right category, so it reaches the right person. Use a short, fixed list and train the desk on it.
| Category | Example at Contoso | Routed to |
|---|---|---|
| How-to or training | User cannot find how to change a delivery date | Key user; work instruction update |
| Authorization | Billing clerk missing a role for a billing app | Security team |
| Data | Customer has the wrong payment terms after migration | Data owner and migration team |
| Configuration defect | Wrong pricing condition for a fleet contract | Functional expert |
| Development defect | Custom output form missing a field | Developer |
| Interface or job | Carrier interface messages failing | Integration or technical expert |
| Product defect | Standard function misbehaves as documented in an SAP note | SAP Support incident |
| Change request | New report wanted for plant managers | Backlog for after hypercare |
Triage categories and routing. Change requests are logged but not built during hypercare unless critical.
Priorities tied to business impact
Define priorities by business impact, not by who shouts loudest. A common scheme has four levels. Align the top level with SAP's own definition when you raise incidents to SAP: SAP's guidance reserves "very high" for very serious consequences for core business processes, such as a production system that is down or a go-live that cannot be completed, with no workaround.
The daily rhythm
- 07:30 stand-upNew P1 and P2 tickets, overnight monitoring alerts, KPI gaps
- Assign and fixOwners named; workarounds communicated by key users
- Fix pathUrgent but controlled change: fix, test, approve, transport
- 16:00 stand-upStatus of open P1 and P2; fixes ready for release
- Daily reportTicket trend by category and priority; KPIs; risks
Fixes need control, even in a hurry
During hypercare, fixes should follow an expedited but controlled change and test path. Each fix is tested in the quality system, retested by the person who raised it where possible, checked for side effects by a short regression of the affected process, approved, and transported in a planned release slot. A fix applied directly in production without testing turns one incident into two.
What would you do?
On day three, billing for several fleet customers fails with a pricing error. Billing is the last step before cash. The consultant has a fix and wants to change the condition record directly in production now. What should Priya recommend?
- Monitor business KPIs, processes, interfaces, jobs and system health, not only tickets.
- Set up SAP Cloud ALM for Operations monitoring and alerts before go-live.
- Triage into a fixed set of categories and route accordingly; log change requests for later.
- Set priorities by business impact and workaround; fixes follow an expedited but controlled path.
Check your understanding · not graded
Lesson check
1. Which monitoring layer most directly shows whether the business is running?
2. During hypercare, fixes should:
3. A user asks for a new plant manager report during hypercare. How should it be triaged?
4. Which issue fits the highest priority?