Heads up, posted this on the Omnissa Community forums (https://community.omnissa.com/forums/topic/71425-windows-updates-causing-issues-for-horizon-connection-servers-nov-2025/), but wanted to put it in here as well... Just waned to give everyone a heads up: We've been noticing with a number of our customers that Windows Updates have been causing issues with servers running Horizon Connection servers. Symptoms: • Horizon Connection Server services fail to start • Horizon Connection Server services started, but Horizon does not function • "Unable to find entity" error banner when viewing desktop pools • "Unable to find entity" error banner when modifying desktop pool settings (edit also fails) In some cases, Horizon servers completely go down, in other cases services function, but errors are seen on the Admin UI ("Unable to find entity"). For customers we managed, we perfromed the following to resolve: 1. After performing Windows Updates, restart servers as you would normally 2. Wait 5-15 minutes post-restart, and gracefully restart the servers again 3. Wait 5-15 minutes post-restart, and gracefully restart the servers a 3rd time After restarting the servers gracefully 3-4 times, everything comes back up and works as expected. PLEASE NOTE: If you have multiple connection servers, stagger the restarts to avoid a race condition. Do not perform the restarts on all connection servers at the same time. I'd recommend performing the actions individually on a single server, completing remediation, verifying services are functioning and up (including web), and then proceeding to the next server.

We've been running into this. The connection servers get Windows Updates spaced 2 hours apart, so there are no simultaneous restarts. Restarting the entire server doesn't usually seem to help (but sometimes does - maybe we just need multiple attempts). Restarting just the Horizon Connection Server service after the system is already up has worked reliably so far (though it takes quite a bit for everything to come up). We were debating changing the service to "delayed start" to see if it helps. Haven't done a deep log dive yet or opened a ticket, but it's been on the list. Have you talked to Omnissa or found any known root cause? We're running Horizon 2503 on Server 2016. It also happens in our test environment running 2503.1, so it's not resolved in that patch. Maybe related, but we've also noticed that since upgrading to 2503, the servers idle at around 5x the CPU utilization that they did prior to 2503.

Omnissa just said they aren't aware of any issues wide-scale unfortunately.

Dug into the heavy CPU usage (with copilot's help) and it looks like it's spinning cycles on something SAML related, burning one core on it at all times. Wonder if that's a factor in the startup issues or a separate, unrelated issue. Are your customers using SAML? 2503?

In the environments I've seen this with, SAML was not being used.

Just noticed https://kb.omnissa.com/s/article/6001200 updated yesterday. at least for high cpu usage in tomcat when SAML is enabled.