Now accepting guest editorial from SAP, Oracle, Workday and ServiceNow practitioners.Pitch a guide
The Sign-Off Table

Lesson 4 of 5

Retest

8 min

Listen to this lesson

0:00 / 6:54

A defect is not finished when the developer says it is fixed. It is finished when the person who found it has run the test again and seen it pass. Cloud ALM's defect workflow is built around that idea, and the test lead's job is to make sure it is followed.

The retest loop

  1. In ProgressThe processor (developer or consultant) takes the defect and works on it
  2. Fix transportedThe fix is moved to the quality system, often via a feature
  3. Retest RequiredThe processor sets this status and makes the original reporter Responsible
  4. RetestThe reporter starts a new run of the test case in Test Execution
  5. ClosedIf the run passes, the reporter closes the defect; if not, it goes back with details
SAP's recommended defect workflow, from fix to closure.
The Sign-Off Table · thesignofftable.com/academy/sap-alm-204

The handover at Retest Required is the critical moment. The processor must set the original reporter as Responsible, because that is who gets notified. The reporter then retests in Test Execution, which means a new run in the same plan. The new run becomes the latest run and sets the test case's status in the plan; the failed run stays in history as evidence of the defect.

Is the fix really in the system?

The most common reason a retest fails is not a bad fix but a fix that has not arrived. Before retesting, confirm that the transport is imported into the quality system. In Cloud ALM, a defect can be linked to the feature that carries the fix, and the feature shows its transports and deployment status. At Contoso, Priya asks developers to put the feature or transport number in the defect before setting Retest Required.

  1. Is the feature or transport number recorded on the defect?Yes → next question · No → developer adds it before setting Retest Required
  2. Is the transport imported into client 100 of the quality system?Yes → next question · No → wait; retesting now will fail for the wrong reason
  3. Is the original reporter set as Responsible?Yes → reporter retests with a new run in the same plan · No → set it, so the right person is notified
Three checks before a retest, so a failed retest means a bad fix, not a fix that has not arrived.
The Sign-Off Table · thesignofftable.com/academy/sap-alm-204

Retest scope: the fix and its neighbors

Retest the action that failed, but also think about what else the fix touched. A change to account determination for plant 1720 billing can affect other billing types, credit memos and intercompany billing. Agree with the developer which related test cases to rerun. For larger fixes late in a cycle, add a short regression set to the plan.

Fix touchesRetestAlso rerun
Pricing condition for fleet customersThe failed order test caseQuotation, credit memo and rebate test cases for the same customers
Account determination for billing in plant 1720The failed billing test caseOther billing types and intercompany billing for 1720
Output form for invoicesThe failed output actionInvoice reprint and email output
Authorization role for UAT testersThe blocked actionA negative test that unauthorized users are still blocked

Choosing retest scope around a Contoso fix.

Retests also need time in the plan. If the cycle ends the day fixes arrive, retests either do not happen or are rushed. Keep the last days of each cycle for retests and agree a cut-off for fixes that will be retested in this cycle; later fixes go into the next cycle's plan.

  1. Days 1 to 10Execution and fixingTesters run the plan; developers fix defects as they arrive
  2. Day 11Fix cut-offFixes delivered after this go into the next cycle's plan
  3. Days 11 to 14Retest windowReporters retest fixes and their neighbors; short regression set for larger fixes
  4. Day 15Cycle exitEvery closed defect backed by a passed retest; postponed defects listed with owners
Keeping the last days of a cycle for retests, with a fix cut-off agreed up front.
The Sign-Off Table · thesignofftable.com/academy/sap-alm-204

Postponed defects

Not every defect gets fixed before go-live. A defect set to Postponed is a business decision, not a technical one. Record who accepted it, the workaround, and the date by which it will be fixed. Postponed defects belong in the go-live readiness pack, because they are known risks the business is choosing to carry.

What would you do?

A Contoso developer fixes a pricing defect, sets it to Closed, and moves on. The original tester never reran the test. Two days later Maria finds the same error in UAT. What process change would prevent this?

Key points
  • Workflow: In Progress, fix, Retest Required with the reporter Responsible, retest run, Closed.
  • A retest is a new run in the same plan; the failed run stays as evidence.
  • Confirm the fix is imported in the quality system before retesting; link the feature or transport.
  • Retest neighbors of the fix; Postponed defects are business decisions recorded for go-live readiness.

Check your understanding · not graded

Lesson check

  1. 1. After fixing a defect, which status should the processor set?

  2. 2. Who should be Responsible when a defect is Retest Required?

  3. 3. How is a retest recorded in Test Execution?

  4. 4. What does a Postponed defect represent?