Update to current ESXI 8.0.3 not possible with upgrade Manager 5.2

[ICPMuenchen]

Hey,

we would like to update our 2 Hosts with the current Upgrade Manager to the latest available Vmware ESXi Update  Q8A57-11207_hpe-esxi8.0u3c-24414501-803.0.0.11.8.0.6-offline-bundle.zip. Only ESXi Update, cause we are already on 5.2.

This is not possible because we get stuck in Pre-Upgrade Validation Tests due to DRS status test failed:

There are only 2 VMs left on the Host: OVC and vCLS. Is there an issue with the current Upgrade Manager? 

HPE Support told us, two possible workarounds would be enabling Retreat Mode to disable vCLS VM or to install it manually via esxcli, but why is it not working with upgrade manager?

Regards

Peter

[support_s]

System recommended content:

1. Gen10/Gen10 Plus Service Pack for ProLiant 2024.11.00.00 Release Notes

2. HPE customized image VMware-ESXi-6.5U3 and 6.7U3 Release notes for VMware vSphere upgrade/update on HPE Hyper Converged 250 and HPE Hyper Converged 380 System

Please click on "Thumbs Up/Kudo" icon to give a "Kudo".

Thank you for being a HPE valuable community member.

[Joshua610]

Thank you for your query - Allow me a moment to investigate & I will reach out by Monday "End of Day" with some information pertaining to your situation - thank you. 

[Joshua610]

This vCLS - what datastore is it on? 

I am researching several items/directions related to your query since it falls on vmWare as well as Simplivity. 

With Simplivity we require vm's to have HA and to be able to relocate during the process of upgrades due to OVC's having to power down. 

Upgrade Guide:

https://support.hpe.com/hpesc/public/docDisplay?docId=sd00005177en_us&docLocale=en_US

I will research further to see if we have any update in future releases. Though the provided work-around is a known workaround. 

Allow me a couple of more days of research for you. Thank you for reaching out. 

[ICPMuenchen]

tried to shut down the vCLS manually, but then the update failed again. support didnt found any issue too.
nevertheless i was able to install the update via esxcli.

thx anyway for your help

[Joshua610]

Good day, 

Thank you for allowing me time to do further research in your direction. I was able to confirm that UM does work/function with vCLS vms living within the environment.

I was able to work with our research team as well as the previous case owner and would like to see if you would allow myself to clone your previous case and let us troubleshoot further to see why your environment isn't allowing this tasks to complete? 

Please let me know if this action would work for you? 

I will monitor for your reply - thank you for reaching out. 

[MGolloher]

I encountered the same issue when I had DRS disabled during patching.

To get past this, I had to enable DRS on the cluster before it would proceed.  

I always liked moved things around manually and take it slow for peace of mind during patch time.

[SIMPLYNICO]

Hello, Same probleme for me.
Should I put the esx in maintenance mode befor starting the update ?

[ICPMuenchen]

thats not working either.

when you shutdown the ovc controller and put the host in maintenance mode, the upgrade manager gives an error, inventory not possible and you cannot proceed.

[Joshua610]

Good day, 

Thank you for allowing me to investigate this further. Whenever I am working with a client on a live update I always have them run the following two commands (1 via PowerShell, where UM is running & 1 via the OVC being upgraded) so that we may identify failures before the UM displays them. 

Via PowerShell -- 

Get-Content “C:\Users\<yourUser>\AppData\Local\UpgradeManager\orchestrator-<date>.log” –wait

Via OVC --
tail -f /var/log/svt-upgrade.log

With @ICPMuenchen's situation I was able to locate the following within the Orchestrator Logs pointing to DRS<important information retracted>: 

{"name":"vSphere DRS Status Validation","id":"xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxx","summary":"The vSphere DRS status test failed.","details":["This host has 1 virtual machines running on it. Enable the vSphere DRS setting for the cluster (license required). Or, power off or migrate the guest virtual machines (excluding the Virtual Controller) on the host."],"state":"FAILED",

Without DRS (license required) updating the ESXi (maintenance mode required) will need to be done manually. I am providing the information within the Upgrade Guide reflecting DRS: 

Here is the link as well for further review: 

https://support.hpe.com/hpesc/public/docDisplay?docId=sd00005177en_us&docLocale=en_US

Please let me know if I can be of further assistance in this direction and if possible feel free to open a case with support for we would love to assist, thank you. 

[ICPMuenchen]

@Joshua610 

but thats not possible, because when you place the host in maintenance mode, UM stops, because it cannot do an inventory anymore. hence the host needs to be in normal operating mode. you either cannot move the vcls machine because it is nailed to the esx host. the only thing would be enabling retreat mode of vcls or shutdown vcls manually to proceed. i didnt try the retreat mode, but the manual shutdown of vcls did not help, because the update failed here too. maybe there was an issue with our host, because in the first attempt i was not able to update the host manually via esxcli, we got a warning that additional bootbank and storage plugins need to be removed first, after a second attempt few days later it worked without any issues.

nevertheless we will try with the next update of esxi when its available. for the moment we are fine.

[FabrizioDV]

Hello

I have encountered this situation too with only the OVC and the VCLS VM living on the host

my solution was the forced shutdown of the VCLS one moment before the pre-requisite check

so the upgrade manager do not find the vcls alive

so the prerequisite will not fail and the procedure will continue

[Joshua610]

Happy Friday, 

With investigating it is stated that without DRS Enabled (License required) one will need to upgrade Esxi Host Manually via ESXCLI. Not with the Upgrade Manager. 

Here is information from HPE regarding Upgrade Manager:

https://support.hpe.com/hpesc/public/docDisplay?docId=sd00005177en_us&docLocale=en_US

The 2 important commands to monitor Upgrade Process & to Identify errors: 

Via PowerShell --
Get-Content “C:\Users\<yourUser>\AppData\Local\UpgradeManager\orchestrator-<date>.log” –wait

Via OVC --
tail -f /var/log/svt-upgrade.log 

Here is some more information from Broadcom/vmWare in regards to their product the vCLS: 

vSphere Cluster Services (vCLS) in vSphere 7.0 Update 1 and newer versions
https://knowledge.broadcom.com/external/article/312147

Introduction to the vSphere Clustering Service (vCLS) [VMware Application Acceleration]
chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://www.vmware.com/docs/vmw-introduction-vsphere-clustering-service-vcls

Embedded vSphere Cluster Services Deep Dive
https://blogs.vmware.com/cloud-foundation/2024/07/17/embedded-vsphere-cluster-services-deep-dive/

Embedded vSphere Cluster Services Overview
https://blogs.vmware.com/cloud-foundation/2024/07/17/embedded-vsphere-cluster-services-overview/

Disable vCLS on a Cluster via Retreat Mode
https://knowledge.broadcom.com/external/article/316514/how-to-disable-vcls-on-a-cluster-via-ret.html

As a recap - If a user has vCLS within their environment and wish to use UM then proper DRS configuration is required otherwise do not use UM and perform the Host upgrade manually via ESXCLI. 

[Joshua610]

Please let me know if there is anything I can do to further assist & as always if needed feel free to reach out to our support team for we love to help. 

Have a great weekend all and I will follow up Monday. 

[Joshua610]

Yes 

[Joshua610]

Upgrade manually without UM. UM requires DRS to be enabled. Or you can disable vCLS ( I believe this works as well as manually) 

Disable vCLS on a Cluster via Retreat Mode
https://knowledge.broadcom.com/external/article/316514/how-to-disable-vcls-on-a-cluster-via-ret.html

[Joshua610]

Yes, UM will not work without DRS enabled - Leaving 2 options 

1 - Manually update without UM via ESXiCLI

2 - Disable vCLS: 

Disable vCLS on a Cluster via Retreat Mode
https://knowledge.broadcom.com/external/article/316514/how-to-disable-vcls-on-a-cluster-via-ret.html

I am here to help so please do not hesitate to reachout with any questions/concerns - thank you. 

[NZamp]

Hello to all,

i'm just activating the Maintenance Mode on the affected ESXi. Then the VCLS will be moved and after that i cancel the Maintenace Mode and the UM can run and Updates can be done.

Hopes this helps.

Best regards,

Nick

[support_s]

Hello,

Let us know if you were able to resolve the issue.

If you are satisfied with the answers then kindly click the "Accept As Solution" button for the most helpful response so that it is beneficial to all community members.

Please click on "Thumbs Up/Kudo" icon to give a "Kudo".