What is Cloud Manager?
Cloud Manager is Adobe's CI/CD platform for AEM as a Cloud Service. It handles everything from code builds and quality checks to environment provisioning and deployments. If you're migrating from AEM 6.5 or AMS, Cloud Manager replaces your custom Jenkins pipelines, deployment scripts, and manual package installations.
While Cloud Manager simplifies many operational tasks, enterprise teams often encounter friction points—especially in multi-tenant architectures or when migrating from highly customized AMS setups.
Challenge #1: Unpredictable Deployment Times
One of the most frustrating aspects of Cloud Manager is deployment time variability. Identical code can take anywhere from 20 minutes to over an hour to deploy, with no clear explanation.
The Problem
Cloud Manager runs on shared infrastructure. Your deployment time depends on:
- Queue position—how many other deployments are running
- Code complexity and number of modules
- Asset processing requirements
- Environment tier (dev vs stage vs prod)
Worse, you cannot cancel a deployment mid-process. Once it starts, you wait.
Workarounds
- Schedule deployments during off-peak hours — Early morning or late evening often has shorter queues
- Use dev environments for rapid iteration — Save stage/prod for validated builds
- Batch your changes — Avoid triggering multiple small deployments
- Set realistic expectations — Plan for 45-60 minute windows, not 20 minutes
Check status.adobe.com before critical deployments. Platform issues can significantly extend build times.
Challenge #2: Single Pipeline Architecture
Cloud Manager uses a single pipeline per program. For organizations with multiple tenants, brands, or teams sharing an AEM instance, this creates coordination nightmares.
The Problem
In AMS or on-premise AEM, teams could deploy independently. Each tenant had their own deployment schedule. With Cloud Manager:
- All code deploys together in one pipeline
- One team's failure blocks everyone
- Release windows must be coordinated across all teams
- Independent releases are not possible without separate AEM programs
Impact on Release Velocity
Organizations that previously deployed twice per week often find themselves dropping to twice per month after migrating to Cloud Manager. The coordination overhead and shared pipeline create bottlenecks that significantly slow release velocity.
Workarounds
- Separate AEM programs — If budgets allow, give independent teams their own programs
- Trunk-based development — Keep everyone on main branch with feature flags
- Release trains — Establish fixed release windows (e.g., every two weeks)
- Modular architecture — Isolate tenant code to minimize cross-impact
Challenge #3: Code Quality Gates
Cloud Manager enforces code quality checks via SonarQube, including security scans, code coverage, and Adobe-specific rules. While valuable, these checks run late in the pipeline—often after you've already waited 30+ minutes for the build.
The Problem
Code that passes all local checks can still fail Cloud Manager's quality gates:
- SonarQube rules differ between local and Cloud Manager
- Adobe-specific checks only run in Cloud Manager
- Failures block the entire pipeline for all tenants
- Feedback loop is slow—you wait 30+ minutes to learn your code failed
Shift Quality Left
The solution is to run quality checks before pushing to Cloud Manager:
# Run SonarQube locally before pushing
mvn clean verify sonar:sonar \
-Dsonar.projectKey=my-aem-project \
-Dsonar.host.url=http://localhost:9000
# Or integrate into your CI pipeline (GitLab, Jenkins, etc.)
# before mirroring to Cloud Manager's Git repo
Recommended Quality Pipeline
- Pre-commit hooks — Lint and format checks
- Local SonarQube — Run the same rules Cloud Manager uses
- CI pipeline checks — Full quality scan before Cloud Manager
- Cloud Manager — Final validation (should rarely fail if you did steps 1-3)
Adobe publishes their SonarQube rule set. Configure your local SonarQube to use identical rules so there are no surprises in Cloud Manager.
Challenge #4: Pipeline Scalability
As your codebase grows—more modules, more tenants, more assets—Cloud Manager build times increase. Organizations with 20+ submodules often worry about hitting scalability limits.
The Problem
Scalability concerns include:
- Build times increase linearly (or worse) with module count
- Large DAM repositories slow down deployments
- No official guidance on maximum module/repository sizes
- Performance degradation as tenant count grows
Scalability Best Practices
- Modularize intelligently — Don't create unnecessary submodules
- Use sparse checkouts — Only clone what you need during builds
- Optimize Maven builds — Parallel builds, incremental compilation
- Separate content from code — Keep large content packages out of code deployments
- Monitor build metrics — Track build times over time to catch degradation early
Challenge #5: Stage Environment Constraints
Cloud Manager provides a single stage environment that serves as both QA and pre-production. For organizations with multiple teams, this creates significant bottlenecks.
The Problem
Stage environment constraints include:
- Only one stage environment per program
- All teams must share the same stage
- Teams going to production block others from using stage
- No parallel testing—sequential access only
- Integration testing requires coordination across teams
Stage Management Strategies
- Release windows — Define clear windows for stage access
- Feature flags — Deploy code to stage disabled, enable for specific testing
- Lower environment testing — Maximize testing in dev before stage
- Automated regression — Run full test suites automatically to minimize stage time
Teams not going to production may need to "back out" of stage during release windows. Build this into your release process—it's a significant change from AMS practices.
Challenge #6: Rollback Is Not Blue-Green
If you're coming from AMS with blue-green deployments, Cloud Manager's rollback feature will feel like a step backward.
| Aspect | AMS Blue-Green | Cloud Manager Rollback |
|---|---|---|
| Rollback time | Seconds (DNS switch) | 30-90 minutes |
| Mechanism | Traffic routing | Code restore from backup |
| Data state | Both environments live | Single environment |
| Availability | Always available | 30-day retention |
Safer Release Strategies
- Feature flags — Deploy disabled, enable gradually, disable if issues arise
- Canary deployments — Test with subset of traffic before full rollout
- Thorough staging — Catch issues before production
- Automated testing — Validate deployments immediately post-release
Challenge #7: Versioning Complexity
Cloud Manager appends unique build numbers to your versions, which can complicate release tracking and troubleshooting.
The Problem
Your version strategy might be:
# Development
1.0.0-SNAPSHOT
# Release
1.0.0
But Cloud Manager produces:
# What you see in Cloud Manager
1.0.0.2024010112345
This makes it harder to correlate what's deployed with what's in your Git repository.
Versioning Best Practices
- Tag releases in Git — Use Git tags to mark what was deployed
- Include commit SHA — Reference Git commits in release notes
- Document build numbers — Track Cloud Manager build numbers alongside versions
- Automate tracking — Script the correlation between Git and Cloud Manager
Challenge #8: Content Migration with CTT
The Content Transfer Tool (CTT) is essential for migrating content to AEM Cloud, but it comes with its own set of challenges.
Common CTT Issues
- Forms migration requires separate utilities
- Large repositories can take days to transfer
- Version retention requires careful planning
- Support is fragmented across multiple Adobe teams
For detailed CTT guidance, see our Content Transfer Tool guide.
Challenge #9: VPN and Network Integration
Enterprise organizations often require VPN connectivity for secure integration with backend systems. This adds complexity to the Cloud Manager architecture.
Network Considerations
- VPN high availability — Ensure redundant connections for production
- Fastly CDN limitations — VPN-enabled scenarios have different CDN behaviors
- Log forwarding — CDN logs may require different handling with VPN
- Security model — Balance zero-trust aspirations with regulatory requirements
Summary: Cloud Manager Success Factors
Despite its challenges, Cloud Manager is a capable platform. Success comes from:
- Adjust expectations — It's not AMS. Release processes will change.
- Shift quality left — Run checks before Cloud Manager, not just in it.
- Coordinate teams — Single pipeline requires organizational alignment.
- Plan for time — Build in buffer for unpredictable deployment times.
- Use feature flags — Compensate for slower rollbacks with safer releases.
- Engage Adobe early — Surface blockers before they become critical.
Cloud Manager trades flexibility for managed simplicity. The organizations that succeed are those that adapt their processes to Cloud Manager's model rather than fighting against it.
Related Topics
Struggling with Cloud Manager?
Our team has navigated complex enterprise migrations to AEM Cloud. We can help you optimize your pipeline and avoid common pitfalls.
Get Expert Help