Creating a Vcenter cluster.

[brunomgr]

Hi Guys

I need some help, i'm trying to install a cluster to do some tests, and basically what i'm trying to do is to use two esxi and put them into a cluster. Simple thing right? But not so easy.

So i have install 2 Esxi without any problem. Next i create a cluster, and for now my conclusions are:

I run the cluster config, and as soon i add one of the hosts the vCls start and stop, saying the message deleted " error the task was canceled by a user".

I'm getting this for about several weeks now and i'm getting a lit bit tired of this. Today i have tried a different approach, the difference was on adding the hosts to the vcenter. For mistake i didn't use the name and use the ip of the esxi. As soon i had it to the cluster, everything worked well. 

Then i add the second one, and again everything was successful, so the question is, why i can't use the full fqdn name? When adding a host  to vcenter using the fqdn, works very well without an issue. 

Both names can be resolved in LAN without an issue.

Any idea?

Thank you

[VRO123]

Remove the ESXi Host from vCenter Once removed, ssh to ESXi and update the FQDN hostname.

To rename a VMware ESXi host:

Log in as root to the console of the ESX host.
Using a text editor, change the name and domain name, if applicable, of the host in these files:
 
/etc/hosts
/etc/sysconfig/network
 
Run this command:

esxcfg-advcfg -s ESXi_FQDN /Misc/hostname
where ESXi_FQDN is the new FQDN hostname for the ESX host.
Reboot the ESX host.
Join the ESX host to VirtualCenter/vCenter Server and clusters.

[brunomgr]

Hi

Thanks for the help, but the issue remains.

When start the deploy of the vCLS image i got the same error, The task was canceled. It stops at 3%.

Again i try with the ip and works. Any more ideas. 

thank you

[kenobi79]

Hi 

verify the entry DNS of the ESXi on DNS Server 

Verify if vcenter resolve the FQDN of esxi (connect in ssh to vcenter server and ping the FQDN of esxi)

If vcenter don't resolve verify the vcenter network configuration: go on https://vcenter_FQDN:5480 with root credentials --> Networking --> edit Networking settings and add the dns servers

alternatively insert the esxi ip and FQDN in vcenter file host

[ptarnawski]

Check: 

• Incorrect DNS configuration. Make sure that the DNS servers on the ESXi hosts and vCenter Server are configured correctly and that the FQDNs for the ESXi hosts can be resolved correctly. You can test this by running the nslookup command on the ESXi hosts and vCenter Server.
• Firewall blocking. Make sure that the firewall on the ESXi hosts and vCenter Server is not blocking the ports that are required for cluster communication. The following ports must be open:
  • TCP 443
  • UDP 443
  • UDP 53
• Certificate issue. Make sure that the SSL certificates on the ESXi hosts and vCenter Server are valid and trusted. You can check the certificate status by connecting to the ESXi hosts and vCenter Server using a web browser.
• vCenter Server bug. There have been some known bugs in vCenter Server that can cause problems with adding ESXi hosts to clusters using FQDNs. If you have checked all of the above and are still having problems, you can try upgrading to a newer version of vCenter Server or searching for a known bug that matches your issu

[brunomgr]

Hi

After 3 days of debugging, i'm still having the same issues. 

the 2 esxi can resolve all the fqdn names of both esxi and the vcenter using ssh / nslookup.

The vcenter fqdn name can be resolved from etheir hosts. Even with this configuration, i put all the names in the hosts file of each esxi and vcenter. 

Now the status is, one of the server, the deploy of the vCLS went successfully and the node was added to the cluster.

When i try to add the second node teh deploy of the vCls does not went successfully. The issue remains.

Transcript from Recent tasks.

Deploy OVF template
vCLS-654dd369-cb32-4451-b325-d3a72a309414
The task was canceled by a user.

[ptarnawski]

Hey @brunomgr ,

Sorry, I forgot to ask but why on Earth do you want to spin vCLS VM? 

[kenobi79]

hi 

from vcenter can you resolve the esxi hosts?

[brunomgr]

Yes I can, I can resolve all hosts from either the vcenter or any esxi.

[brunomgr]

I don't want to spin vCLS VM. The process take place when I add a second node of the cluster. 

[kenobi79]

Hi 

is possible to have the error? Or the Error Logs?

I attach some links on VCLs

https://kb.vmware.com/s/article/91891

https://core.vmware.com/blog/troubleshooting-vsphere-cluster-services-vcls-vms-retreat-mode

https://docs.vmware.com/en/VMware-vSphere/7.0/com.vmware.vsphere.resmgmt.doc/GUID-96BD6016-4BE7-4B1C...

https://kb.vmware.com/s/article/83076

https://kb.vmware.com/s/article/83984

PS: consider the possibility to open a case for support

[brunomgr]

Hi

The error that i'm getting is this one, as soon i add the second node to the cluster

Deploy OVF template
vCLS-654dd369-cb32-4451-b325-d3a72a309414
The task was canceled by a user.

[kenobi79]

hi 

did you see this kb?

https://kb.vmware.com/s/article/2117310

[ptarnawski]

Hey @brunomgr ,

Maybe I do not understand what are you trying to do. 

You have a vCenter, you added 2 ESXi hosts. You want to put those hosts into the cluster. IS that correct? 

Yo mentioned that there is an error related to vCLM ? 

If so disable DRS and HA for the time being, make your configuration and then enable DRS and HA 

[brunomgr]

Hi

That's like that. A center with two esxi and I want to add them to the the cluster, nothing more.

The only error I got is related with vCLS that cannot deploy a ovf image when I add the second node. 

I have DRS and HA disabled. I just do the config of adding two esxi to the cluster, nothing more.

[brunomgr]

Hi

Anymore ideas, i already do all of your recommendations, but still got the same error.

Deploy OVF template
vCLS-00a72eff-3045-4c92-b665-9ef9c918692e
The task was canceled by a user.

[kenobi79]

hi 

I think that is better to open a Service Request to Support

[ptarnawski]

If DRS and HA are disabled vLCM should not be deployed, 

Create another cluster, make sure both are disables and try again with new cluster

[brunomgr]

You're right, vCLS  shouldn't be deployed, but I have all services disabled so there is no reason why. And that's the problem. Why is it deployed and how I remove it?