Skip to main content

CRE-2025-0179

N8N Workflow Silent Data Loss During ExecutionCritical
Impact: 9/10
Mitigation: 7/10

CRE-2025-0179View on GitHub

Description

N8N workflow automation platform experiences critical silent data loss where items \ndisappear between workflow nodes without generating error messages. This high-severity \nissue affects long-running workflows (60-115+ minutes) and can cause workflows to \nrandomly cancel mid-execution, leading to incomplete processing and data integrity \nproblems. Items silently vanish between nodes, with different item counts across \nthe workflow pipeline, making the issue particularly dangerous for production systems \nthat rely on complete data processing.\n

Mitigation

- **Implement workflow item counting checks** - Add validation nodes between critical \n processing steps to verify item counts match expected values\n- **Enable comprehensive execution logging** - Set N8N_LOG_LEVEL to debug and \n EXECUTIONS_DATA_SAVE_ON_SUCCESS to 'all' to capture detailed execution data\n- **Add workflow timeout monitoring** - Monitor executions that cancel around 21-23 \n minute mark and implement retry mechanisms for failed workflows\n- **Implement data integrity validation** - Add checksum or validation steps at \n workflow start/end to detect silent data loss\n- **Use error handling workflows** - Configure error workflows to capture and log \n execution failures, even when main workflow fails silently\n- **Monitor execution metrics** - Set up alerting on workflow completion rates and \n item processing inconsistencies\n- **Consider workflow segmentation** - Break long workflows into smaller, more \n manageable chunks to reduce exposure to the data loss issue\n

References