Update of vIDB to 9.0.2 via Fleet Management fails with Error LCMVMSP10037

Products

• VMware SDDC Manager / VCF Installer
• VCF Operations

Issue/Introduction

When updating a vIDB cluster to version 9.0.2 via the Fleet Management appliance, the process fails during the patch download phase. The binary transfer is interrupted prematurely, resulting in an incomplete file.

Symptoms:

• The update task in the UI fails with error code LCMVMSP10037.

• Manual curl download of the vmsp.tar bundle fails exactly after 3072 MB with: curl: (18) transfer closed with outstanding read data remaining.

  curl -k https://#############/repo/productPatchRepo/patches/vidb/9.0.2.0/vmsp.tar -o vmsp.tar
  
    % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
  
                                   Dload  Upload   Total   Spent    Left  Speed
  
  100 3077M    0 3077M    0     0  2268k      0 --:--:--  0:23:09 --:--:-- 3095k
  
  curl: (18) transfer closed with outstanding read data remaining

• Log entries in /var/log/vrlcm/vmware_vrlcm.log on the Lifecycle Manager (vRLCM) show:

ERROR vrlcm[189734] [pool-3-thread-42] [c.v.v.l.v.p.t.VmspPatchApplicationPlatformTask]  -- -- BUNDLE [vmsp-platform/vmsp] STATUS: Failed --
failed to download: unexpected EOF

INFO vrlcm[189734] [pool-3-thread-42] [c.v.v.l.p.a.s.Task]  -- Injecting task failure event. Error Code : 'LCMVMSP10037', Retry : 'true', Causing Properties : '{ CAUSE :: fqdn === productPatchFileName === vmspPatchFileName === primaryVip === productId === version === rollbackPatch ===  }'
com.vmware.vrealize.lcm.vmsp.common.exception.VmspPatchingException: -- BUNDLE [vmsp-platform/vmsp] STATUS: Failed --
failed to download: unexpected EOF

        at com.vmware.vrealize.lcm.vmsp.plugin.tasks.VmspPatchApplicationPlatformTask.execute(VmspPatchApplicationPlatformTask.java:81) [vmlcm-vmspplugin-core-9.0.2.0-SNAPSHOT.jar!/:?]
        at com.vmware.vrealize.lcm.automata.core.TaskThread.run(TaskThread.java:62) [vmlcm-engineservice-core-9.0.2.0-SNAPSHOT.jar!/:?]
        at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source) [?:?]
        at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source) [?:?]
        at java.base/java.lang.Thread.run(Unknown Source) [?:?]

Environment

 

VMware Cloud Foundation (VCF)
Fleet Management / vIDB Cluster

Target Version: 9.0.2

 

Cause

The Nginx reverse proxy on the Fleet Management appliance has a default proxy_max_temp_file_size limit of 3072m. The 9.0.2 vmsp.tar patch bundle exceeds this size (approximately 3077 MB).
When the download reaches the 3GB threshold, Nginx stops buffering to disk and closes the connection, leading to an "Unexpected EOF".

Resolution

To resolve this issue, increase the Nginx temporary file buffer limit on the Fleet Management appliance.

1. Log in to the Fleet Management appliance via SSH as root.

2. Navigate to the Nginx configuration directory: cd /etc/nginx/

3. Back up the current configuration: cp nginx.conf nginx.conf.bak

4. Edit nginx.conf with vi

5. Locate the location /repo/ block.

6. Change the proxy_max_temp_file_size value from 3072m to 4096m.

   location /repo/ {
     proxy_pass http://vrlcm-server;
     proxy_max_temp_file_size 4096m;
     proxy_temp_path /data/temp/;
   }
   

7. Save the changes and exit the editor.

8. Validate the Nginx configuration syntax: nginx -t

9. Reload the Nginx service to apply the changes: nginx -s reload

10. Retry the vIDB update from the Fleet Management UI.