"one or more hosts that need disk format upgrade" but all hosts/disks are showing same & latest

[MillardJK]

Does anyone know how to fix it when the cluster configuration status shows "one or more hosts that need disk format upgrade", listing less than all the hosts (in my case, it's 3 out of a 5-node cluster), yet all the disk groups indicate they're at the same Disk Format Version in both the Web Client UI and rvc (in my case, v3 for a hybrid 6.0U2 cluster)?

[DMFDMinister]

Just chiming in to say I have the exact same problem:

Virtual SAN cluster Converged Team Cluster has one or more hosts that need disk format upgrade: dellesx1.starwars.local, dellesx2.starwars.local, dellesx3.starwars.local, dellesx4.starwars.local. For more detailed information of Virtual SAN upgrade, please see the 'Virtual SAN upgrade procedure' section in the documentation

I have 20 All-Flash disks. Each disc is on Disk Version 3.0. Disks with outdated version: 0 of 20.

-vSAN 6.2

- Version 6.0.0 Build 3617395

VCSA - 6.0.0.20000

PSC - 6.0.0.10200

Hosts have latest patches as provided by Update Manager as of yesterday. This is a vSAN lab for my team, so nothing critical running on it at the moment.

The only piece of hardware not certified for vSAN in the lab is a Fusion IO SLC High IOPS adapter for each host, but they're working perfectly. Hosts are R710s.

[admin]

Greetings!

This is a known issue. Please try restarting vCenter server services.

For Windows based vCenter - Stopping, starting, or restarting VMware vCenter Server 6.0 services (2109881) | VMware KB‌‌

For vCenter Server Appliance - Stopping, starting, or restarting VMware vCenter Server Appliance 6.0 services (2109887) | VMware KB‌

If this issue does not get resolved, please open a case with VMware Technical Support. They would most likely need to do a cleanup on internal database.

Hope this helps you with the next actions.

_________________________

Was your question answered correctly? If so, please remember to mark your question as answered when you get the correct answer and award points to the person providing the answer. This helps others searching for a similar issue.

Cheers!

-Shivam

[MillardJK]

I've tried the vCenter Restart tactic to no avail; looks like an SR is in my future...

[admin]

As a last try, can you please retest the VSAN Health by following path - Virtual SAN Cluster > Monitor > Virtual SAN > Health > Retest

If this does not go well, please quote Bug# 1563082 and 1641574 while opening SR.

[darcidinovmw]

If you do end up opening an SR, please let us know the outcome.

[vpradeep01]

Hello,

Good day !

Assuming you just rebooted/restarted the vCenter server, could you please restart the vsan health service on all the ESXI hosts too? This will not impact your running VM's.

/etc/init.d/vmware-vsan-health restart

Did you upgrade the vCenter server and ESXI host before the ondisk upgrade? If yes which one was first upgraded. Was it vC and then ESXI?

Thanks

[millardjkONITS]

/etc/init.d/vmware-vsan-health restart

My hosts are running 6.0.0.4192238, and that command doesn't exist. It *does* exist on the vCenter server, which has been restarted more than once.

Did you upgrade the vCenter server and ESXI host before the ondisk upgrade? If yes which one was first upgraded. Was it vC and then ESXI?

Here's how it went: I installed ESXi 6.0U2 and used one of the SSDs to 'bootstrap' the VCSA; for some reason, the download I got--thinking it was also 6.0U2--was actually 6.0GA. So when I got around to checking on the VSAN health after building out the cluster, I saw that it wasn't present (the health service, that is). After researching things a bit, I discovered what had happened and did an in-place upgrade of vCenter to 6.0U2; the hosts were (correctly) building VSAN using the v3 file system, but the vcsa was totally confused on the version; after the update to U2, the file version warning persisted on the 3 hosts.

[vpradeep01]

Hello,

I'm sorry for that.

Ok, I think its just a warning in the vSAN Cluster. You should be be able to "suppress warning" on the summary page and then clear/acknowledge if any alarms. This would have triggered when you upgraded your ESXi version first instead of vCenter server.

Since the host was at the latest level 6.0 U2 and vC was still at 6.0 GA, the warning might have triggered which is expected.

Although the vCenter server -> Manage -> Virtual SAN General reports as 3.0, you can confirm this from rvc too.

Run these commands from the rvc and get me the output please?

vsan.obj_status_report .

vsan.disks_stats .

Fire them under the cluster level.

[MSDS]

We are also having this exact same issue after we upgraded out of order. Our health checks come back PASSED and I have restart vCenter and the Health services. The Alarm remains. Our vSAN Tech said it's a vCenter issue more so dealing with HA. I'm having issues finding where to suppress the alarm. Our tech mentioned having a vCenter tech clear this via the database. Anyone has experience with navigating the VCSA 6.0U2 db?