Lesson 3 of 5
Deployment orchestration
10 min
A Contoso release is never just ABAP. Release 1 contains configuration and code transports for S/4HANA, Cloud Integration flows for customer and supplier interfaces, and a few manual steps such as maintaining a background job. Orchestration means these arrive in production in the right order, at the right time, together, and that you can prove it.
One release, several transport mechanisms
SAP describes deployment orchestration as the synchronous deployment of multiple features by release timebox, made consistent by the landscape information and by the deployment plan assigned to the project. In practice: assign every feature to the release, check that its transports are released, run transport checks across the bundle, then deploy the features together.
Order matters
Within a deployment, Cloud ALM deploys released transports and ignores modifiable ones, and imports first those transports that are furthest from production. Across mechanisms, you decide the order. An integration flow that calls a new custom API in S/4HANA must arrive after the ABAP feature that delivers the API. Write these dependencies on the features and check them in the deployment review.
- FreezeRelease 1 scope frozen; no new features assigned
- CheckTransport checks across the bundle: downgrade and cross-reference conflicts resolved
- ApproveLuis confirms every feature is Ready for Production
- Deploy ABAPMass deploy the S/4HANA features to CMP
- Deploy integrationDeploy cTMS features once the ABAP objects they call are in place
- Manual stepsCutover tasks run in sequence from Luis’s plan
- VerifySmoke tests in production; features confirmed Deployed
Automating the lower tiers
Not every deployment needs a meeting. The Deployment Scheduler lets you schedule transport deployments by system role (quality assurance, pre-production, production). Contoso schedules nightly imports into the quality assurance role during test cycles so testers always work on the latest build, and keeps production imports manual and approved. Landscapes with a pre-production tier often schedule that tier too, aligned with regression runs.
| Target | Trigger | Who | Evidence |
|---|---|---|---|
| Quality assurance (CMQ) | Nightly schedule for features In Implementation or In Testing | Automated; Sam monitors failures | Import logs; Feature Traceability shows status |
| Pre-production, if present | Scheduled before each regression cycle | Automated; regression lead confirms | Regression run linked to the release |
| Production (CMP) | Release window or approved urgent change | Luis approves; Sam executes | Feature history, import logs, smoke test results |
Deployment triggers by tier at Contoso.
When an import goes wrong
ABAP transports cannot simply be "undone" in production: importing an older version means another transport, and data created in between does not go away. So decide in advance, per release, how you respond to a failed deployment. Most teams prefer fix forward with an urgent feature when the problem is small and understood, and keep a fallback (switching off the new function, reverting configuration with a prepared transport, or a documented manual workaround) for larger failures. For a go-live, the cutover plan's go/no-go and fallback points cover this.
- Is the failure isolated to one feature with a known cause?Yes → fix forward with an urgent feature · No → next question
- Does it stop a critical process such as shipping or invoicing?Yes → invoke the prepared fallback and inform the process owner · No → next question
- Can the business work around it until the next release?Yes → record the workaround and plan the fix · No → urgent path today
What would you do?
During the Release 1 production deployment, the ABAP features import successfully, but one Cloud Integration flow fails because a value mapping it needs was never transported. What should Luis do first?
- Orchestration ties ABAP, integration and manual steps to one release timebox.
- Cloud ALM deploys released transports, furthest from production first; you define cross-mechanism dependencies.
- Run transport checks across the whole bundle before the production deployment.
- Automate lower-tier imports with the Deployment Scheduler; keep production approved and attended.
Check your understanding · not graded
Lesson check
1. How does SAP describe deployment orchestration in Cloud ALM?
2. An integration flow calls a new API delivered by an ABAP feature. What order should production deployment follow?
3. Which tier does Contoso keep manual and approved?
4. Why rehearse the production deployment in a copy of production?