vLCM Sync Updates fails with 'A depot is inaccessible or has invalid contents' in VCF 9.1
search cancel

vLCM Sync Updates fails with 'A depot is inaccessible or has invalid contents' in VCF 9.1

book

Article ID: 445052

calendar_today

Updated On:

Products

VMware SDDC Manager / VCF Installer

Issue/Introduction

  • Log into the vSphere UI. Select hamburger icon > Lifecycle Manager > ACTIONS > Updates > Sync Updates.
  • This error is observed:

Sync updates fails - vCenter error when trying to sync updates. A general system error occurred: "A depot is inaccessible or has invalid contents. Make sure an official depot source is used and verify connection to the depot."

  • In the vCenter log - /var/log/vmware/vmware-updatemgr/vum-server/imageservice.log - this error is observed:

Two baseimages share the same releaseID: ESXi:9.1.0-0100.25433460, but have different attributes.

Environment

VCF 9.1.0.0

Cause

  • The online correct base image and an imported base image are clashing.
  • This is caused by creating a 9.1 ESXi image while the vCenter was running 9.0.

Resolution

(1) Perform a cleanup of the vLCM depot:

  1. Take a snapshot, without memory, of the vCenter.
  2. SSH to the vCenter, log in as root.
  3. Launch the dcli:
    • dcli +i +show
  4. List the offline depots (log in as administrator@vsphere.local when prompted):
    • esx settings depots offline list
  5. Identify the ID for the problematic depot (e.g., OFFLINE_BUNDLE_xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx.zip - the one with the 9.1 ESXi image, imported when the vCenter was still running 9.0).
  6. Delete the problematic depot:
    • esx settings depots offline delete-task --depot XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX --force true
  7. Delete all orphan images in the vLCM image library, so that the only images that are left are the ones that are assigned to a cluster, via the vSphere UI > Lifecycle Manager > Image Library


(2) Perform a cleanup of UMDS content inside of SDDC Manager:

  1. Take a snapshot, without memory, of the SDDC Manager.
  2. SSH to the SDDC Manager as vcf, then su to root.
  3. Remove the contents of the UMDS:
    • rm -rf /nfs/vmware/vcf/nfs-nount/umds/patch-store/hostupdate
  4. Verify removal:
    • ls -l /nfs/vmware/vcf/nfs-mount/umds/patch-store/hostupdate/vmw/

To confirm that the solution recommendation has resolved the issue, log into the vSphere UI > Lifecycle Manager > ACTIONS > Updates > Sync Updates

This workflow is expected to succeed now, when it failed previously.