Raudi Tracker

Re: Permission Audit of Vcenter Groups and Members

Raudi — Wed, 22 Nov 2023 12:35:20 GMT

The same error messages i had too, you need to disable the appliance shell.

Before using the script enter this on the VCSA:

chsh -s /bin/bash root

and when the script is finished this command:

chsh -s /bin/appliancesh root

The error message from the script is exact the same when you enter the command in the appliance shell...

Re: VM Storage Privileges missing in 8.0U1

Raudi — Sat, 24 Jun 2023 18:23:02 GMT

Thank you very much for that detailed answer!

Re: NVIDIA Tesla P4 Passthrough Not capable

Raudi — Fri, 23 Jun 2023 14:44:33 GMT

I think you must decide if you want use "PCI Passthrough" with that card or you want to use vGPU with the drivers...

Try to uninstall the NVIDIA drivers from the Host, perhaps then it is possible to use the card for PCI Passthrough.

I'm using this card here in vGPU Mode, then the card can serve 4 VM's at one time. For that mode you need to use the NVIDIA drivers on the host.

Re: VMware Tools problems on a updated ESXi host (6.7->7.0)

Raudi — Fri, 23 Jun 2023 11:39:33 GMT

I think i'm getting it...

The /tools folder is a Ram-Disk Cached folder. Because this host is still using SD cards.

The new installed hosts have bigger SSD's for the boot device, here is no Cache Ram-Disk.

The behavior is described here:

Installing and upgrading the latest version of VMware Tools on existing hosts (2129825)

Re: VMware Tools problems on a updated ESXi host (6.7->7.0)

Raudi — Fri, 23 Jun 2023 11:25:19 GMT

Okay, this is interesting...

After a reboot of the host, the content of /locker/packages/vmtoolsRepo is copied to /tools and the link of productLocker is changed back to /tools.

I noticed this, because i tryed to delete the /tools content after changing the link, but i was only able to delete the content, not the folder. Then after a host reboot the folder is filled again and with the same new version.

And on a different host i looked to the /tools folder before and afer the reboot. Before there are the old VMware Tools and after the reboot the folder content was synced to the current version.

Why is the behavior on a clean installed and updated host different? Why must the complete tools folder exist twice on the updated host?

VMware Tools problems on a updated ESXi host (6.7->7.0)

Raudi — Fri, 23 Jun 2023 10:52:48 GMT

Hi,

we have noticed a problem with the installation of the current VMware Tools on a ESXi 7.0 host which was updated from 6.7.

On the updated host the tools are in a /tools folder, the /productLocker links to /tools.

This is the content of /tools/vmtols folder:

-rw------- 1 root root 130664448 Mar 6 10:47 windows.iso

And on the fresh installed 7.0 host the /productLocker is pointing to /locker/packages/vmtoolsRepo

This is a different location, and this path exists on the updated host too, with this content:

-rw-r--r-- 1 root root 130775040 Jun 16 13:40 windows.iso

So the install routine of the VMware Tools are installing the Tools to a wrong location on the updated host, or during update from 6.7 to 7.0 the Tools location wasn't configured correctly.

I have now executed this both commands:

rm productLocker
ln -snf /locker/packages/vmtoolsTepo /productLocker

After a host reboot the VM's are showing that a Tools update is available.

The question is now, what can i do, that without a host reboot, the new VMware Tools are detected by the VM's?

And is this a konwn issue, or should i open here a support request?

Kins regards

Stefan

Re: VM Storage Privileges missing in 8.0U1

Raudi — Thu, 22 Jun 2023 11:02:54 GMT

Hi Ben,

but what must be selected if i need:

Profile-driven storage.Profile-driven storage update
Profile-driven storage.Profile-driven storage view

In the past there was this 2 privileges, now i see there 7... Wich one of this 7 are the correct ones, which refelfts the two above?

If i should guess, i would say:

Update VM storage policies
View VM storage policies

But i need this for shure... On the other hand, i told our Veeam specialist already, the best will be to open a case at Veeam, they must say what is now the correct one.

Kind regards

Stefan

VM Storage Privileges missing in 8.0U1

Raudi — Tue, 20 Jun 2023 12:18:34 GMT

Hi,

someone just asked me where to find the Profile Driven Storage privileges documented here:

VM Storage Policies Privileges (vmware.com)

On a vCenter 7.0 we found them, but in vCenter 8.0U1 we don't find this. Because we want to create a role for a Veeam user...

Do we missed something?

Kind regards
Stefan

Re: Problems after updating VCSA 8.0 -> 8.0U1

Raudi — Sat, 06 May 2023 07:38:35 GMT

Very strange, i made again the U1 update after my restore and have done a 2 restarts of the vcenter after that within the last days, but i can't reproduce this issue.

The problem last time, i had la few people here, they want to start their view session, so i need a running vCenter.

And the main reason for this post was, to chek if someone else had a similar experience.

But this problems seems to have no one else and after the second update i havn't this issue again.

Because this it must bave been a one time issue.

Problems after updating VCSA 8.0 -> 8.0U1

Raudi — Mon, 24 Apr 2023 06:37:26 GMT

Hi,

on thursday evening i updated my VCSA to 8.0U1, i had no problems... Yesterday, 3 days later, i had some maintenance and had to shutdown my complete environment. After startup of the VCSA i was able to login, but there was an error message that i don't remember, i was able to finish the startup of the VM's.

Perhaps 6 hours later i got a notice that the connection from a PCoIP Zero Cleint don't work. 1,5 hours before it was possible.

So i tryed to login to the VCSA, and after login i wasn't able to see something i have no rights to view was the message. And there was an error that he can't access the /sdk page.

When i try to access the /sdk page i get "no healthy upstream" the same when i go to the root of the page.

On SSH i was can see that the services vmware-content-library, vmware-vpxd and a few more are not running. So i was able to start some but some not. They can't be started, because there was no access to the /sdk page.

I made a complete reboot, but the VCSA don't came back. And i didn't found something that was wrong.

All certificates are valid up to next year, i have a fixed IP and DNS is working.

Then i have now deleted that VCSA and restored it from a storage snapshot of thursday morning, some hours before the U1 update.

Now the VCSA is running normal and i will try the update again.

Has someone seen something like that?

Regards
Stefan

Re: Questions regarding VAAI Snapshot Offload -> Setting VM options

Raudi — Fri, 24 Mar 2023 21:31:36 GMT

Thank you, perfect idea to post that there.

I used in the past this script to enable that feature on all my VM's:

Connect-VIServer -Server vcenter.server.local -user administrator@vsphere.local -password PASSWORD
$vms = get-vm -Location "Datacenter-Name"
foreach ($vm in $vms) {
New-AdvancedSetting -Entity $vm -Name snapshot.alwaysAllowNative -Value TRUE -Confirm:$false -Force:$true
}

Mounting a VMDK of quiesced snapshot shows a uninitialized disk

Raudi — Thu, 22 Dec 2022 09:25:22 GMT

Hi,

a Windows Server 2016 VM on a most recent vSphere 7 host. The VM is runing, i create a quiesced snapshot and mount the original VMDK as a "Independent - Nonpersistent" disk to a different VM.

If all is running as expected i must see a additional "offline" disk in the disk management.

But not in this case, here i see only a "not initialized" disk.

If we do a not quiesced snapshot and monting this disk to a different VM, all is working as expected.

So, doing a quiesced snapshot seems to result in a corrupt disk.

We are seeing manny events in the event log, like described here:

Creating a quiesced snapshot of a Windows virtual machine generates Event IDs 50, 57, 137, 140, 157, or 12289 (2006849) (vmware.com)

But this didn't help in this case.

And this behavior we have only with a few VM's, there are many other VM's where i can add a disk from a quiesced snapshot without problems.

Has someone a hint or has seen something similar?

Regards
Stefan

Re: VM's still selected after vMotion...

Raudi — Mon, 19 Dec 2022 09:34:53 GMT

Hi, that will be great, thank you!

VM's still selected after vMotion...

Raudi — Sat, 17 Dec 2022 14:06:10 GMT

Hello,

today i had a bad experience in my current vSphere 8.0 environment...

To install the latest patches i was on one host, on the tab with all VM's on that host i selected all running VM's except 3 with local devices.

This VM's i moved with drag/drop to a other host.

After this was finished i selected the 3 other VM's to shutdown the VM's.

But what i missed, at the status line at the bottom of the VM list window, still shows that more than 20 VM's are selected. So i have shutdown manny of the VM's i just moved. Not all, because the vCenter was one of them.

This behavior can't be correct and is very dangerous!

Has someone seen this too? Is this a known issue and on the list to be fixed? Or must i open here a SR? The problem, we have only 5 SR in a year, and it is bad to waste this for bugs...

Regards
Stefan

Re: No standby possibe after U3?

Raudi — Thu, 15 Dec 2022 16:15:48 GMT

As i wrote, do you have VMware HA enabled? If yes disable this for a test, because shutting down one host in a 2 node HA cluster hasn't enoug resources in case of a failure.

Re: No standby possibe after U3?

Raudi — Thu, 15 Dec 2022 15:27:25 GMT

I do this every time manually.

I switch the 2nd host on only when i install updates. All the other time the host is off to save power.

And perhaps in a 2 node cluster that don't work because of VMware HA to have the failover resources?

Re: Microsoft Cluster with physical RDM - Problem with 2nd node

Raudi — Fri, 23 Sep 2022 11:04:52 GMT

O.k. the vCenter told me that he was updated the last time in may and i configured the last cluster in june. There we migrated from FC to iSCSI. I removed the FC mapping files from the VM and created new ones for the iSCSI devices with no problems...

Re: Microsoft Cluster with physical RDM - Problem with 2nd node

Raudi — Fri, 23 Sep 2022 10:59:14 GMT

Yes...

And a few moth ago i made 2 clusters with no problems, i must take a closer look if we installed a vCenter update before or after this.

Because this i'm totally confused, normally i know how to do that.

Re: Microsoft Cluster with physical RDM - Problem with 2nd node

Raudi — Fri, 23 Sep 2022 10:43:32 GMT

I have done so, look what i have written under the first screenshot...

Re: Microsoft Cluster with physical RDM - Problem with 2nd node

Raudi — Fri, 23 Sep 2022 09:53:40 GMT

Pherhaps a "Bug" in vCenter? After writing that i had the idea to made a try locally on the ESXi Host, not in the vCenter, to add this existing disk:

I can save it and then in the vCenter i can switch the "Sharing" to "Multi-writer" and save it with no problems:

Hmm....