Host compliance check for hyperconverged cluster configuration

[mihkel]

I'm building a 3 node vsan standard ESA cluster (ThinkAgile VX3331 Node)

Vsan skyline health points out one issue:

---

Description:Check if the configuration of each host is compliant with the hyperconverged cluster configuration. The hosts should all be attached to the VDS with the same adapters, the adapters should be present and on the correct distributed port groups, and the NTP server and lockdown mode should be configured according to specifications

Risk if no action taken:At least one host in the vSAN cluster is not compliant with the hyperconverged cluster configuration, which may cause managability issue or even lead to usability issue for the vSAN cluster.

---

It implies that node1 and node3 are not configured.

But when I go to "Cluster -> Configure -> Configuration -> QuickStart", then configure is grayed out. See attachment.

 

vsan cluster info from each node:

 

node1:

esxcli vsan cluster get
Cluster Information
Enabled: true
Current Local Time: 2023-05-23T06:56:25Z
Local Node UUID: 642b0f7d-7356-a999-14dd-84160c45ad04
Local Node Type: NORMAL
Local Node State: AGENT
Local Node Health State: HEALTHY
Sub-Cluster Master UUID: 646725dc-4dde-eaaf-eed9-84160c4549ce
Sub-Cluster Backup UUID: 64676141-1b7b-26d7-888a-84160c44c034
Sub-Cluster UUID: 5248d338-6263-e073-6655-de31ae6a51ba
Sub-Cluster Membership Entry Revision: 4
Sub-Cluster Member Count: 3
Sub-Cluster Member UUIDs: 646725dc-4dde-eaaf-eed9-84160c4549ce, 64676141-1b7b-26d7-888a-84160c44c034, 642b0f7d-7356-a999-14dd-84160c45ad04
Sub-Cluster Member HostNames: node2, node3, node1
Sub-Cluster Membership UUID: 44666b64-4820-0458-b7fd-84160c4549ce
Unicast Mode Enabled: true
Maintenance Mode State: OFF
Config Generation: 0b8b6bbf-0e86-4978-8729-d189a64019d5 40 2023-05-23T04:04:02.210
Mode: REGULAR
vSAN ESA Enabled: true


node2:

esxcli vsan cluster get
Cluster Information
Enabled: true
Current Local Time: 2023-05-23T06:56:28Z
Local Node UUID: 646725dc-4dde-eaaf-eed9-84160c4549ce
Local Node Type: NORMAL
Local Node State: MASTER
Local Node Health State: HEALTHY
Sub-Cluster Master UUID: 646725dc-4dde-eaaf-eed9-84160c4549ce
Sub-Cluster Backup UUID: 64676141-1b7b-26d7-888a-84160c44c034
Sub-Cluster UUID: 5248d338-6263-e073-6655-de31ae6a51ba
Sub-Cluster Membership Entry Revision: 4
Sub-Cluster Member Count: 3
Sub-Cluster Member UUIDs: 646725dc-4dde-eaaf-eed9-84160c4549ce, 64676141-1b7b-26d7-888a-84160c44c034, 642b0f7d-7356-a999-14dd-84160c45ad04
Sub-Cluster Member HostNames: node2, node3, node1
Sub-Cluster Membership UUID: 44666b64-4820-0458-b7fd-84160c4549ce
Unicast Mode Enabled: true
Maintenance Mode State: OFF
Config Generation: 0b8b6bbf-0e86-4978-8729-d189a64019d5 40 2023-05-23T04:04:02.215
Mode: REGULAR
vSAN ESA Enabled: true


node3:

esxcli vsan cluster get
Cluster Information
Enabled: true
Current Local Time: 2023-05-23T06:56:30Z
Local Node UUID: 64676141-1b7b-26d7-888a-84160c44c034
Local Node Type: NORMAL
Local Node State: BACKUP
Local Node Health State: HEALTHY
Sub-Cluster Master UUID: 646725dc-4dde-eaaf-eed9-84160c4549ce
Sub-Cluster Backup UUID: 64676141-1b7b-26d7-888a-84160c44c034
Sub-Cluster UUID: 5248d338-6263-e073-6655-de31ae6a51ba
Sub-Cluster Membership Entry Revision: 4
Sub-Cluster Member Count: 3
Sub-Cluster Member UUIDs: 646725dc-4dde-eaaf-eed9-84160c4549ce, 64676141-1b7b-26d7-888a-84160c44c034, 642b0f7d-7356-a999-14dd-84160c45ad04
Sub-Cluster Member HostNames: node2, node3, node1
Sub-Cluster Membership UUID: 44666b64-4820-0458-b7fd-84160c4549ce
Unicast Mode Enabled: true
Maintenance Mode State: OFF
Config Generation: 0b8b6bbf-0e86-4978-8729-d189a64019d5 40 2023-05-23T04:04:02.213
Mode: REGULAR
vSAN ESA Enabled: true

 

vcenter version: 8.0.1 Build: 21560480

ESXi version: VMware ESXi, 8.0.1, 21495797

vsan traffic is on vmk2 and configured on all 3 nodes.

How to proceed with this issue? Although this is new vsan cluster and running only few VMs for testing purposes (including vcenter), I'd like to avoid rebuilding it from scratch.

[TheBobkin]

, Have you tried pressing the 'Add' button, selecting the nodes then applying the QS configuration to them?

 

QS is completely optional - your cluster can be configured completely fine without it - your cluster looks to be fully-formed (as you evidently configured it at least partially manually instead of via QS) and this this isn't a big concern.

[mihkel]

Hey.

 

Yes I have tried 'Add'. Nothing I can do there, since all hosts I have are already in cluster. It just bothers me that according to UI, everything is not green, although vsan cluster is formed and functional.

 

Any way to tell UI, that mark all hosts as configured? Since UI thinks that there are some issues, I also cannot configure reserves and this is a must in particular case.

 

I have also tried to completely remove one host from cluster according to this https://kb.vmware.com/s/article/2148975

but when I add it back via QS, then I'm back to square one - 2 host are not configured.

[jconduit7]

Did you ever figure this out, I have the same issue?

[jason_dvorak]

Also running into this issue.  Has a resolution been found for this?

[mihkel]

Nope. I have not found a solution. But cluster itself is running just fine

[ajacobsendk]

Did you change the VDS configuration after you set it up ? I found out that using quick start, if you change uplink nic after the initial config, i ran into this issue, i found no way to get around it.

Br.

[jason_dvorak]

Our situation was a little interesting because of NSX integration with our hosts.  We had an issue that was only solved by moving the hosts one at a time into a new temporary cluster, re-configuring NSX on them one at a time there, then moving them back to the production cluster.  So while the vDS configuration itself didn't change on the hosts, the cluster did treat them as if they had not yet been configured using quick start.

[mihkel]

 I really do not remember any more, but it might be that I changed VDS configuration.

I tried to completely remove hosts one by one from cluster(https://kb.vmware.com/s/article/2148975), but for me it did not solve anything