setup an example windows machine

proxmox-ha-win-vm.md

Setup HA Windows VM

this example uses windows server 2022. Windows 11 will be basically the same

It is beyond this gist to be a deep tutorial.

this gist is part of this series

Prep

1. Download your windows ISO of choice (i am using windows 2022 Data Center for this example.)
2. Download the lates virtio.iso too
3. go to Datacenter > pve1 > expand child nodes in left pane > local (pve1) (i don't know why i find this unintutive)
4. click ISO Images
5. click upload and path to your ISO and click upload - you will need to do this on every node you want to create VMs on (ISOs are not replicated by default - but see this is you want to replicate across nodes)

Create VM

1. on the node you uploaded the click the create VM in the upper right corner
2. on the General page you need to specify a name and the click next
3. on the OS page

• set the ISO to boot, select your storage and then select the iso image from the drop down
• set guest OS to windows and the latest date
• click next

4. For the system page it is imperative you pick the pool you created ealier for EFI and TPM storage as follows:

5. Set the disk to virtio block storage for max perf
6. it is also important to select the ceph pool when creating disks for the VM as follows (note my personal preference to use write through for the cache policy - i don't care about loosing reads, i do care about loosing writes; the discard setting is outside the scope of this gist.

do not start the VM at this point.

Attach and configure Virtio Block Storage driver during OS install

1. navigate to Datacenter > pve1 > vmnode > hardware
2. click add
3. choose cd / dvd drive
4. set as SATA2
5. choose your storage and the virtio ISO Image and click ok.
6. now boot the VM and press key to boot from DVD
7. when it ask where you want to install OS it will ve blank - click load driver
8. click browse
9. navigate to the virtio CD drive (e.g D: or whatver letter it has)
10. expanf the file structure and select d:\viostor\2k22\amd64 and click ok
11. you should see the Red Hat Virtio SCSI controller listed - if so click next.
12. now you can select the unallocated space and continue with install
13. when you can login to the OS install everything on the virtio CD using the istall application in the root of the CD
14. shutdown machine and change network type to virtio and GPU to VirtIO-GPU (and unmount virtio cd and install dvd) and reboot

Configure for HA

1. in the gui select Datacenter > HA
2. click add in the resources pane
3. select ClusterGroup1 (this was created in an ealier gist in this gist sequence)

Test Live Migration

1. select the VM from pve1 node treet in the GUI
2. start the VM if it is not already started and wait fo bootstorm activity to subside
3. click migrate in the top right corner of the UI
4. select another node say pve3 and select migrate - note it should say online like the following:

Your vM should migrate in a minute or so at maximum with no errors.

Test Failed Node

1. make sure your VM is started on pve3
2. pull the power on node pve3
3. watch it fail over in about 3 to 4 minutes

@scyto
I came across this content in the "help" section of the Hard Drive creation screen. Not sure if it's just more recent info or not, but it appears VirtIO SCSI is the most modern driver that provides high performance. Is there something more to consider that pushed you toward VirtIO Block?

"A SCSI controller of type VirtIO SCSI single and enabling the IO Thread setting for the attached disks is recommended if you aim for performance. This is the default for newly created Linux VMs since Proxmox VE 7.3. Each disk will have its own VirtIO SCSI controller, and QEMU will handle the disks IO in a dedicated thread. Linux distributions have support for this controller since 2012, and FreeBSD since 2014. For Windows OSes, you need to provide an extra ISO containing the drivers during the installation.

The VirtIO Block controller, often just called VirtIO or virtio-blk, is an older type of paravirtualized controller. It has been superseded by the VirtIO SCSI Controller, in terms of features."

You will also find in latest proxmox you can no longer create vms with block. Can't remember why I chose it, it will have been because I read somewhere it was better at perf. Either way is moot now it's virtioscsi we all have to use. Get Outlook for iOS<https://aka.ms/o0ukef>

…

________________________________ From: rlabusiness ***@***.***> Sent: Friday, April 12, 2024 1:30:22 AM To: rlabusiness ***@***.***> Cc: Mention ***@***.***> Subject: Re: scyto/proxmox-ha-win-vm.md @rlabusiness commented on this gist.

________________________________ @scyto<https://github.com/scyto> I came across this content in the "help" section of the Hard Drive creation screen. Not sure if it's just more recent info or not, but it appears VirtIO SCSI is the most modern driver that provides high performance. Is there something more to consider that pushed you toward VirtIO Block? "A SCSI controller of type VirtIO SCSI single and enabling the IO Thread setting for the attached disks is recommended if you aim for performance. This is the default for newly created Linux VMs since Proxmox VE 7.3. Each disk will have its own VirtIO SCSI controller, and QEMU will handle the disks IO in a dedicated thread. Linux distributions have support for this controller since 2012, and FreeBSD since 2014. For Windows OSes, you need to provide an extra ISO containing the drivers during the installation. The VirtIO Block controller, often just called VirtIO or virtio-blk, is an older type of paravirtualized controller. It has been superseded by the VirtIO SCSI Controller, in terms of features." — Reply to this email directly, view it on GitHub<https://gist.github.com/scyto/054c153defe3acbbe3a61d6fe08704d4#gistcomment-5021000> or unsubscribe<https://github.com/notifications/unsubscribe-auth/ACWXVXAOR5DCD2S6C3JT6N3Y46LR7BFKMF2HI4TJMJ2XIZLTSKBKK5TBNR2WLJDUOJ2WLJDOMFWWLO3UNBZGKYLEL5YGC4TUNFRWS4DBNZ2F6YLDORUXM2LUPGBKK5TBNR2WLJDHNFZXJJDOMFWWLK3UNBZGKYLEL52HS4DFVRZXKYTKMVRXIX3UPFYGLK2HNFZXIQ3PNVWWK3TUUZ2G64DJMNZZDAVEOR4XAZNEM5UXG5FFOZQWY5LFVEYTENBRHA2DEOJXU52HE2LHM5SXFJTDOJSWC5DF>. You are receiving this email because you were mentioned. Triage notifications on the go with GitHub Mobile for iOS<https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android<https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

Oh - that’s good to know, and it makes the decision easy. :-)

Oh - that’s good to know, and it makes the decision easy. :-)

yeah i only realized this last week, at some point i will update my instructions

i was thinking about it more, i think when i originally built these the virtioscsi didn't support discard, i can no longer remember why i thought that was important....

...i can no longer remember why i thought that was important....

If you're referring to not remembering why the support for "discard" is important, I believe it's because that's required for VMs to pass through TRIM commands to the host in order to keep the SSD/NVMe performing at its peak.

If you're referring to not remembering why the support for "discard" is important,

no i am not, i know whty discard is important, i was referring to why i chose the vitio block device and why i chose it - i think I chose it because at the time scsi didn't suport SSD/NVME features, but thats a hazy recollection , either way not relevant as we only have option in UI for vitioscsi now.