Loading
Salesforce now sends email only from verified domains. Read More

Batch Jobs in Stuck State After Application Restart for On-Premise Applications in Mule 4

Publish Date: Sep 30, 2025
Description

When a Mule 4 on-premise application is restarted while a batch job is still running, the job instance may remain in the EXECUTING state.

  • Graceful restart (stop/start or redeploy): In most cases, Mule attempts to resume pending jobs. However, there are scenarios where the batch job does not resume correctly and remains stuck.

  • Abrupt restart / JVM kill: If the JVM is killed or the restart happens abruptly during batch execution, the Object Store used for the batch instance may become corrupted. This prevents the job from resuming and leaves stale entries behind.

  • maxAge behavior: The maxAge setting only clears job instances that are in a Expired state. Jobs left in EXECUTING are not eligible for cleanup and will attempt to resume.

As a result, subsequent scheduled jobs can also get stuck in the Loading phase, since the previous instance is still considered active and blocking execution.

Resolution
  •  

To resolve this issue:
1. Undeploy the application.  
2. Confirm the corresponding folder is deleted under $MULE_HOME/.mule/<app-name>.  
3. Redeploy the application fresh.  

This clears any stale batch or object store state that may have been left during the restart.  

Additional Resources

By default, Batch uses ORDERED_SEQUENTIAL, based on its timestamp, and processes records one by one. Refer Documentation https://docs.mulesoft.com/mule-runtime/latest/batch-reference

Mule 4 Batch Processing Reference](https://docs.mulesoft.com/mule-runtime/latest/batch-reference)  
Salesforce KB: When to Modify the .mule Folder](https://help.salesforce.com/s/articleView?id=001118518&type=1)  

Knowledge Article Number

005224545

 
Loading
Salesforce Help | Article