System Performs Unexpected COLD Start After Zero Downtime Upgrade (ZDU)

Products

• CA Automic Workload Automation - Automation Engine

Issue/Introduction

After performing a Zero Downtime Upgrade (ZDU), the first subsequent reboot of an application server node causes the initial Work Process (WP) to start in COLD start mode. This occurs even if the rest of the system is active and processing tasks on other nodes.

• The system performs an unwanted COLD start despite a normal start being specified.
• Tasks may become stuck in an 'Inconsistent' status.
• Batch processing and automation workflows are disrupted.
• The issue typically triggers during the first planned or unplanned maintenance/reboot following a ZDU completion

Environment

Automic Automation in the process of ZDU from 24.4.x to 24.4.x

Cause

The root cause is a product defect where the MQLS table (specifically the SESSION row) retains the old version information after a Primary Work Process (PWP) switch during the ZDU. When a WP starts after a node reboot, it checks this table; because the version in the database does not match the binary version, the system incorrectly triggers a COLD start

Resolution

This issue has been fixed in Automation Engine 24.4.4.

After you are on this version 24.4.4 or superior, during the PWP switchover the SESSION key in MQLS will be correctly updated to avoid the cold start.

Additional Information

 DE180257

A problem has been fixed where the Automation Engine system performs unwanted COLD-Start on first reboot after a ZDU.