JCP forced traces | Unexpected message version for FH_CONT
search cancel

JCP forced traces | Unexpected message version for FH_CONT

book

Article ID: 443357

calendar_today

Updated On:

Products

Automic SaaS CA Automic Workload Automation - Automation Engine CA Automic One Automation

Issue/Introduction

The Java Communication Processes (JCP) or Worker Processes (WP) generate continuous forced traces, occurring multiple times per second. This activity can cause the JCPs to become overwhelmed, leading to high resource utilization and potential performance degradation of the Automation Engine.

The JCP/WP logs show the following error: 

U00045014 Exception 'java.lang.IllegalStateException: "Unexpected message version for FH_CONT, expected 1, got 3" 
U01002211 Unexpected message version for FH_CONT, expected 1, got 3 
U00003620 Routine 'com.automic.kernel.impl.AgentDispatcher' forces trace because of error.
U00009907 Memory view 'Unexpected version for FH_CONT, received 3, expected 1' (Address='####', Length='####')

Environment

  • Automic Automation Engine: 24.x (SaaS and On-Premise)
  • Agents: Linux OS Agents (seen on 24.3.0)
  • Feature: File Events (EVNT_FILE)

Cause

This is a product defect (DE189662).

The issue is triggered when a file event hit is processed outside of its defined validity window. This commonly occurs in the following scenario:

  1. An event is paused or stopped (e.g., for a maintenance window).
  2. While the event is inactive, the file conditions are met, or a "hit" is registered by the agent.
  3. When the event is resumed or restarted, if the triggered hit falls outside the original validity period, the agent sends a message that the AgentDispatcher in the JCP cannot correctly parse, resulting in an "unexpected message version" exception.

Resolution

argeted to be fixed in a future release of the Automation Engine. Please subscribe to this article (see How to subscribe to Knowledge Base articles) to be updated on the fix status.

Workaround: To stop the loop of forced traces immediately:

  1. Identify the specific RunID mentioned in the JCP traces or Agent logs (e.g., U02000163).
  2. Manually Stop the affected Event task.
  3. Restart the task using Modify -> Go (Release). This clears the invalid message state and resumes normal processing.