DocumentationExperience Cloud KCS

AEM Cloud Manager pipeline fails at deploy step with HELM_RELEASE_FAILED and 503 errors

Last update: October 2, 2026

This article explains how to troubleshoot and resolve deployment failures in Adobe Experience Manager (AEM) Cloud Manager pipelines where the deploy step fails with HELM_RELEASE_FAILED and HTTP 503 errors during health checks.

Description description

Environment

  • Adobe Experience Manager as a Cloud Service
  • Cloud Manager deployment pipeline
  • Publish environment during rollout and startup

Symptoms

  • Pipeline executions consistently fail at the deploy step.
  • The error HELM_RELEASE_FAILED is reported during environment deployment.
  • Health checks for the publish service are marked CRITICAL due to HTTP 503 errors on the /libs/granite/security/currentuser.json endpoint during rollout.
  • After the environment startup, the same endpoint returns HTTP 200, indicating recovery.
  • Pipeline and environment logs show unresolved application dependencies or startup errors.

Cause

Deployment failures are primarily caused by unresolved application dependencies and timing issues during environment startup, resulting in temporary health check failures (HTTP 503) that cause the pipeline to report HELM_RELEASE_FAILED.

Resolution resolution

Follow the steps below to resolve the issue:

  1. Review your application code and confirm that required dependencies are included and compatible with the deployment environment. Check dependencies such as twitter4j, Composum, and ES6 clientlibs if your application uses them.
  2. Review the Cloud Manager pipeline and AEM environment logs for startup errors or unresolved dependencies that could block deployment.
  3. Review the Publish health-check result for /libs/granite/security/currentuser.json during rollout and after Publish startup. Confirm whether it returns HTTP 503 during rollout and HTTP 200 after startup. Note: A later HTTP 200 shows that the endpoint recovered; it doesn’t identify why readiness was delayed. Correlate the health-check timing with the pipeline and AEM startup logs. Don’t change health-check timing unless Adobe confirms a supported, deployment-specific option.
  4. After correcting any identified code, dependency, or configuration issue, rerun the Cloud Manager pipeline and confirm that the deployment completes successfully.
  5. If the pipeline continues to fail, contact Adobe Support. Provide the recent pipeline execution logs, relevant AEM startup logs, the health-check timeline, and details of the code or dependency changes.
recommendation-more-help
experience-cloud-kcs-help-kbarticles