Users experience a critical timing and reporting discrepancy where the Unix Agent prematurely reports a job as successfully finished. Specifically:
Unix/KSH scripts executed via UC4 jobs sporadically terminate in the Automation Engine with Exit Code 0 (Ended Normally) in less than a second.
The AE job report is truncated/incomplete and stops logging prematurely.
Because the AE registers a successful completion, subsequent dependent jobs in the Job Plan trigger prematurely.
Actual Behavior: On the OS level, the executed script is still actively running. It may complete successfully seconds later, or it may fail and exit with a non-zero return code (e.g., Exit 1) which the Automation Engine entirely misses
Note: The issue is intermittent and typically affects a small percentage (~1%) of script executions under heavy production load.
Product: Automic Workload Engine
Component: Unix / Linux Agent
Version: 24.4.x
DE184202
A problem was identified, whereby the exit codes of completing user_service instances could have been used as exit code of normal jobs in case if a pid reuse happened over time between the two processes.
Update to a fix version listed below or a newer version if available.
Fix version:
Component(s): Unix / Linux Agent
Automation.Engine 24.4.4 HF2 - Planned release End of May 2026
Automation Engine 24.4.5 - planned released June of 2026
Automation Engine 26.1.0 - planned released June of 2026
Public Title: Unix Agent not waiting until job ends
Public Description: Unix Agent might consider user_service exit codes for other processes when pid reuse happens