AEM as a Cloud Service: Stage deployment pipeline delays or stuck during deployment
This article explains how to troubleshoot situations where the Adobe Experience Manager (AEM) as a Cloud Service Stage deployment pipeline takes longer than expected or appears stuck during the deployment phase in Cloud Manager.
Description description
Environment
Adobe Experience Manager as a Cloud Service (AEMaaCS)
Symptoms
When deploying to the Stage environment using Cloud Manager in AEM as a Cloud Service, the deployment pipeline remains in the deployment phase for an extended period. Statuses such as waiting for secondary pods to start or progressing are displayed for a long time, even if no recent code changes have been made. Deployment logs show that publish containers are taking longer to become ready, with health checks running continuously until all pods are healthy. Example log entries include generic Sling Model errors such as:
org.apache.sling.scripting.sightly.SightlyException: Identifier <ModelClass> cannot be correctly instantiated by the Use API ...
Caused by: org.apache.sling.models.factory.PostConstructException: Post-construct method has thrown an exception for model class <ModelClass> ...
Caused by: java.lang.NullPointerException: ...
Despite these delays, the deployment may eventually complete successfully once all pods in both the primary and secondary regions become healthy.
Resolution resolution
Follow the steps below to resolve the issue:
- Review the Cloud Manager pipeline execution logs to identify which phase is taking longer (For example: wait-for-env-ready or update-services).
- Check the status of all pods in both the primary and secondary regions using the Cloud Manager interface to ensure they’re progressing towards a healthy state.
- Confirm that publish containers are initializing and passing health checks. Delays will occur if containers require additional time to become ready.
- If you notice custom code errors in the logs (such as NullPointerException in a Sling Model), review and address these errors in your codebase. However, such errors may not always block deployment, if all pods eventually become healthy.
- Wait for the deployment to complete, as the pipeline will proceed once all pods are healthy.
- If the deployment remains stuck for an excessive period or fails, collect relevant logs, and contact Adobe Support for further investigation.
To verify resolution, ensure that the deployment pipeline completes successfully and all pods are reported as healthy in Cloud Manager.