Sean Massey-1

I agree with @Graeme Gordon. Persistent desktop DR is really hard. The big questions with any Horizon DR plan is around the DR plan you'll be designing against and any business or technical requirements that are in place that you will need to meet. What is your "Recovery Time Objective" (RTO - how long it takes to get the environment back in service) and "Recovery Point Objective" (RPO - how much data loss can you tolerate). With Horizon, another factor is business requirements around use case and user expectations. Who are the users? What applications are they using? What are they doing on their desktops? What expectations do the business and the users have and/or what has been communicated to them? These are all questions that come into play when designing for Horizon DR, whether it is persistent or non-persistent. If you can provide a few more details, I can help talk you through some options and considerations.

Do you see any errors on the vCenter side? Have you opened a ticket with support?

Hi Matt, Thanks for the update. This makes sense. Reverting a snapshot in this case would have caused a USN rollback issue. That's one of the drawbacks of using AD LDS/ADAM/LDAP as a replication layer.

Yes, it is possible. Build a Linux template, don’t join it to the domain and deploy it without configuring a domain in the pool settings. You will need some way to manage users on the Linux desktop after deployment.

I forgot to add one thing. I always recommend opening a support ticket proactively when planning for multi-pod Horizon upgrades. This way the ticket is open in case you need it, and you're not wasting time trying to get a ticket opened if you run into issues during the upgrade.

I agree with @robryan. A week between pod upgrades should be fine. So here is the thing on VMware Tools and the Horizon Agent. The Horizon Agent used to have it's own customized version of the SVGA driver, and if you installed or upgraded VMware Tools and the Agent in the wrong order, the customized driver would be overwritten. That stopped being an issue by Horizon 8 2111 when the SVGA driver was removed from the Horizon Agent, and now you no longer need to install or upgrade Tools before the Agent. This is documented in KB2118048, which states the following: You can find the KB here: https://kb.omnissa.com/s/article/2118048

What happens when you hover your mouse over it? Does a window pop up wiht more information?

Generally speaking, yes. But there are a number of corner cases where the details kind of matter, which is why I was asking those questions. One Enterprise Application should be fine in most use cases. In my opinion, the best place to set up MFA, especially Azure MFA, is on the UAGs. The question about internal MFA comes down to "who is required to use MFA internally" and "what is the use case driving it."

You’re right. There is nothing that prohibits it. It’s just a strongly held recommendation across almost the entire EUC space. The automation to support that recommendation is not one of those things that happens overnight. It can take years to get a completely automated build process, especially in complex and app-dense environments. I hope you understand that those of us who have a strong reaction towards in-place upgrades have been burned in the past.

I have not seen a global schema mismatch before. I’m not sure how you would fix that. Typically schema and replication error recovery requires an authoritative restore of the LDAP instance. I’m not sure if there is a process for that in a Cloud Pod environment. I’m sure you’ve looked at it, but I’m going to ask anyway. Have you ruled out any firewall issues at site 2 that might be blocking traffic back to site 1?

There was a VMware KB that covered in-place upgrades for Horizon. Unfortunately, I am having trouble finding it now due to EUC being spun out. If I can locate it again, I will post it.

So a couple of followup questions on this. Has it ever worked with an iPad? Did it work with previous versions of Horizon and something change recently like upgrading to 2406 or upgrading/changing your UAGs? Or is this a completely new thing you're doing and it doesn't work on the iPad? How many users are impacted by this? Is this a test user or a 1-off? Is it happening to specific users connecting to specific machines in a larger pool, or does this happen to anyone trying to use connect to their machines with an iPad? Does it work if you connect to Horizon and that desktop from an iPad on your local network bypassing the UAG?

Hi Lola. I'm not an MSP, but I used to support EUC MSPs when I worked at VMware. So I'm happy to help here.

IIRC, that tool was a fling. It never had official support, and I believe the main developer behind it has since left VMware/Omnissa.

I agree 100% with @StephenWagner7. In-place upgrade is fine for individual physical laptops and desktops. And you may have to do it for some persistent VDI. But you should NOT do an in-place Windows version upgrade if you're building non-persistent VDI or managing your templates. Build a new image on the version of Windows that you're targeting and deploy that instead.