vCLS not start up

[st-ops]

Hello,

we are currently evaluating implementation of ESXi, VCF and vSAN.

The issue I face is at setting up is proper functionality of DRS and cluster-wide-updates.

When checking the events, I can find following notification:

Date Time:

 01/13/2025, 08:53:36

Type:

 Warning

Target:

 vCLS-c6803716-a2d4-8300-852f-a036bcc7a5b7

Description:

vCLS-c6803716-a2d4-8300-852f-a036bcc7a5b7 does not exist on xxxxxxx in xxxxxxxx

Event Type Description:

The virtual machine does not exist on the host with which it is associated

Possible Causes:

The virtual machine was deleted while its host was disconnected from vCenter Server.

Related events:

There are no related events.

I already tried to install the whole system twice, first time I had added ESXi hosts and VCSA with IPs, the 2nd time with DNS names. Both situations with the same result.

I am using distributed switch to set everything up, vMotion and vSAN appear to be working fine.

In the first installation I was up-to-date with everything, currently in the process of updating still (the only way to update hosts is doing manual migrations of VMs, freeing up hosts and then remediating, since automatic migrations - DRS - is not working).

Btw. also there are not DRS recommendations, even if using manual.

I also tried going to retreat mode, waiting for vCLS to disappear, and activating again, without success.

Versions:

ESXi: VMware ESXi, 8.0.3, 24022510

VCSA: 8.0.3.00400

Am I missing something vital? I remember this working without a hitch on previous environments, which were 7.x. I just started working with 8.x recently in the new company.

Thanks a lot.

[Andrea Consalvi]

Hey,

Yeah, I see what you mean. Normally, vCLS should just deploy itself automatically, but if it’s missing and DRS isn’t working, something’s definitely off. If you’ve already reinstalled everything twice and tried both IPs and DNS names, I’d start by checking if the vCLS VMs are actually missing or just stuck.

Jump into vCenter > Cluster > Monitor > vSphere Cluster Services and see if they’re listed. If they’re there but disconnected or not responding, sometimes just removing them forces vCenter to recreate them. Otherwise, you can try disabling and re-enabling vCLS manually to see if that kicks it back into action.

Another thing worth checking is if there’s anything blocking their creation—like resource limits or storage issues. vCLS VMs need a bit of CPU, RAM, and disk space, and if something’s tight, they might not deploy properly. Since you’re on vSAN, it’s also worth making sure that’s fully healthy and not causing issues.

Lastly, DNS. I know you tried both IPs and names, but vCenter really likes having solid name resolution for everything to work smoothly. If you SSH into vCenter, you can do a quick nslookup on your ESXi hosts and see if there’s any delay or failure resolving them. Sometimes just fixing a small DNS hiccup can bring everything back to life.

If none of that helps, it might be worth checking the logs in /var/log/vmware/vpxd.log to see if vCLS is actually trying to deploy and failing for some reason.