Update Manager Service Crashes and creates core.worker.***** dumps

Products

• VMware vCenter Server

Issue/Introduction

• Update Manager service crashes immediately after starting.

• Multiple core dumps for the service are generated in the storage/core partition of vCenter with name:

  • core.updatemgr-worker.###
• Logs in updatemgr-vmon-prestart.log show PostgreSQL routine overloads: WARNING: Routine vc.pgstattuple(text) in database VCDB is overloaded by vc.pgstattuple(regclass)

Environment

VMware vSphere 7.0.x / 8.0.x

Cause

Stale entries or PostgreSQL schema conflicts in the VMware Update Manager (VUM) database.
This naming collision during service startup prevents the database from initializing correctly, causing the service to crash and stop.

Resolution

Resetting the VMware Update Manager (VUM) Database:

1. Connect to the vCenter Server Appliance (VCSA) 7.0 or later via SSH.

2. Enter the BASH shell by running:

   shell

3. Stop the VMware Update Manager service with the following command:

   service-control --stop vmware-updatemgr

4. Reset the Update Manager database, based on your vCenter version:

   • For vCenter Server Appliance 7.0 or later:

     python /usr/lib/vmware-updatemgr/bin/updatemgr-utility.py reset-db

5. Clear the Update Manager patch store by running:

   rm -rf /storage/updatemgr/patch-store/*

6. Restart the Update Manager service:

   service-control --start vmware-updatemgr

Additional Information

For assistance, see Contact Broadcom Support.