VCF 9.1 install issues - VCF Operations and VCF Cloud Proxy using Fleet Lifecycle

[Mark Morsing]

Hello everyone

I'm deploying a greenfield VCF 9.1 and it's gone pretty smooth up until this point.
Deploy and configure the operations appliance - fails at 5/17 with a wall of text, repeating itself over and over.

The Fleet lifecycle 'OVA component installation' task with ID 019e92d0-f852-7ab0-a624-c2bdd3b495c1 failed. Internal errors: Failed to deploy VCF Operations node. Check vCenter tasks for errors

I've looked through the vCenter and all i can see is that it starts deploying the OVF and it cancels the deployment exactly 8 seconds later - it repeats this over and over again until the SDDC installer throws the errors.

I've checked all DNS lookups, regular and reverse.
I've checked NTP - everything is within 0,5 seconds of eachother.

Anyone got any bets on what it could be? I've tried searching through the logs on the SDDC manager / VCF installer but nothing really shows - the reference token just leads me to "check vcenter" and the events + tasks shows deployment + cancellation.
The VPXD log also just shows that the "user" deploys and the "user" cancels...

[Evandro Sardinha]

Hello Mark,

I'm having the same problem. However, I don't see any tasks in vcenter, the VCF Operations deployment was completed, in my case it seems to be a problem with the VCF Cloud Proxy deployment. 

This is my log displayed in the UI:

The Fleet lifecycle 'OVA component installation' task with ID 019e9e62-db25-7be5-b319-04827d971835 failed. Internal errors: Status polling timed out. Review the logs to check operation progress. Increase the timeout value if the operation requires more time., Unexpected server error has occurred., Status polling timed out. Review the logs to check operation progress. Increase the timeout value if the operation requires more time. Status polling timed out.

This is my log from : /var/log/vmware/vcf/domainmanager/domainmanager.log

2026-06-07T03:38:47.770+0000 ERROR [vcf_dm,6a24e712bf2a81196a79a2e0ee8ad95e,ff4d] [c.v.v.v.s.fleetlcm.FleetLcmService,dm-exec-43] Fleet LCM task 019e9e62-db25-7be5-b319-04827d971835 failed with status: FAILED
2026-06-07T03:38:47.771+0000 ERROR [vcf_dm,6a24e712bf2a81196a79a2e0ee8ad95e,ff4d] [c.v.v.v.s.f.FleetLcmTaskExecutor,dm-exec-43] Fleet LCM task for OVA component installation failed, task ID: 019e9e62-db25-7be5-b319-04827d971835, status: FAILED
2026-06-07T03:38:47.783+0000 ERROR [vcf_dm,6a24e712bf2a81196a79a2e0ee8ad95e,ff4d] [c.v.e.s.o.model.error.ErrorFactory,dm-exec-43] [EGL3JG] PUBLIC_FLEET_LCM_TASK_FAILED The Fleet lifecycle {0} task with ID 01
9e9e62-db25-7be5-b319-04827d971835 failed. Internal errors: Status polling timed out.

7-4011-a9d3-7bc584c4516a'] Timeout while waiting for updates. Retrying with version '1'
2026-06-07T03:53:34.774+0000 DEBUG [vcf_dm,6a24e6502f37fa6832d39ae7f06faad1,f28a] [c.v.v.d.notif.impl.pc.PcPollingTask,vcf-ds-2] [Sub Id: '8f180083-b288-4ff5-8350-1c6c5acb3776', Data source id: 'a3cf00b7-8627-4011-a9d3-7bc584c4516a'] Timeout while waiting for updates. Retrying with version '1'

In my case, all the requirements were successfully met: DNS, NTP, etc.

I'm still investigating. If I find anything, I'll share it.

[Evandro Sardinha]

Hello Mark, just wanted to share that I resolved my issue by manually configuring the initial using the New installation (Not Express) option of VCF Operations. After that, the deployment completed successfully.

[idefix90]

Have you perhaps moved vCenter to a different location? For example, to a resource pool? Please check the DomainManager.log file for entries related to OVATOOL.

[Mark Morsing]

Hey @idefix90

Nope everything is completely stock and havn't been touched during the installer.
I havn't been able to find anything related to the OVA deployment neither in the SDDC manager or in the vCenter - despite it cancelling the deployment.
Do you have an idea about what to look for? Then I'll check it out.

[Mark Morsing]

Cool - "Sadly" my logs contain no errors or failures - only cancelled by user.
I also used the express installer so that might be a lead worth investigating.

I'm currently waiting for engineering support to come up with a solution, as they couldn't see a reason for the failure either during troubleshooting.
I'll keep the post updated with findings

[Mandy468]

I had same situation.
Checked - /var/log/vmware/vcf/lcm/lcm-debug.log

The log showed inventory synchronization warnings and errors, including:

Logical inventory not yet loaded after restart
Failed to load hardware support inventory
Failed to get domain inventory

Review - /var/log/vmware/vcf/domainmanager/domainmanager.log

The following error was identified:

Failed to get network profiles type

Actual certificate thumbprint doesn't match provided one for address  'node-vcf-esx01.xxxx.local'

Certificate was correct validate manually.  But Command on SDDC : curl -v https://http://node-vcf-esx01.xxxx.local//sdk

Observed:     SSL certificate OpenSSL verify result:  unable to get local issuer certificate (20)

This suggested a potential certificate trust or inventory synchronization issue between VCF services and the ESXi host. 

POA -       sudo /usr/bin/service-control --restart lcm   and     sudo /usr/bin/service-control --restart domainmanager

instead of Restarting Services,  I just perform SDDC Appliance Reboot,   And later, re-tried the task and completed successfully. 

============================================================================================================

VCF Operations deployment failed due to an apparent ESXi certificate thumbprint validation/inventory synchronization issue within SDDC Manager. A reboot of the SDDC Manager appliance refreshed the internal state and allowed the Fleet Lifecycle deployment to complete successfully. 

[Mark Morsing]

Hello @Mandy468

Thanks for the pointers - sadly none of them are in my logs.
I tried both rebooting and restarting services anyway but none of them solved the issue for me.

I'll post once support finds a solution.

[Andy del Hierro]

So I saw this problem and it was due to out so sync NTP servers. 

[Inter Soldi]

I ran into something very similar during a VCF deployment where the OVF would start and then get cancelled almost immediately with very little useful information in the installer logs. In your case, did you ever check whether there were any hidden vCenter events or resource policy restrictions causing the deployment to abort after those few seconds? Jewellok

[Mark Morsing]

Hello everyone

It all came down top missing firewall port openings.
When you go through the installer and it starts deploying - everything is deployed from the SDDC Manager / Installer, so here we had all the required port openings to/from.

But during the installation, when it deploys the VCF Operations, the deployment shifts away from the installer and into 4 Kubernetes nodes deployed inside the management cluster.
These 4 kubernetes nodes requires the same port openings to/from the ESXi hosts as the SDDC Manager / Installer did - otherwise the install fails.

This behavior is undocumented both public an form the engineers.
The deployment is now finished and two KBs will be published (They're currently under review)

KB443202
KB443201