Version

Argent Advanced Technology 3.1A-1404-A or earlier

Date

Thursday, 19 June 2014

Summary

Occasionally customer may receive system event ‘Task terminated prematurely‘

The following line can also be found in Service Log

If further checking the Relator trace log, the relevant task is actually executed properly

The issue is caused by the timing of recycling of shared Monitoring Engine process

This issue happens most likely in busy environment

The issue is corrected in Argent AT 3.1A-1404-T4

Technical Background

The shared Monitoring Engine process recycles once a while

The maximum running length is controlled by registry

HKLM\Software\Argent\{PRODUCT}\ RULE_ENGINE_MAX_RUN_SECONDS

If the work order file of a task is generated, but not picked up by the Monitoring Engine process yet; and Monitoring Engine happens to recycle at the exact moment, Supervising Engine will see the Monitoring Engine process is terminated, and task has been terminated unexpectedly too

The system event is fired

But actually the restarted Monitoring Engine process will pick up the work order and execute the task properly

The code has been corrected in Argent AT 3.1A-1404-T4 for this scenario

Resolution

Upgrade to Argent AT 3.1A-1404-T4 or later

For customer who cannot upgrade immediately, he can edit registry

HKLM\Software\Argent\{PRODUCT}\ RULE_ENGINE_MAX_RUN_SECONDS to have value ‘0‘

This simply turns off the recycling of shared Monitoring Engine process