Horizon 8 NSX advanced load balancer horizon health check don't work well

[Michael Modro]

We have the problem that the horizon Health check don't work well.
We do the health check of horizon connection servers like discribed (favicon.ico) -> Understand and Troubleshoot Tunnel Connections | Omnissa

In some temporary cases the favicon-check is ok but one of the connection server is not working.
It looks like this from the other connection servers:


On the not working connection server the check looks like this:



But the favcion-Check is healty and so the user get no connection if the load balancer choose the not working connection server.

Do you have another health check for the NSX Advanced load balancer which notice this (maybe a rest api call or somethink like that?

kind regards

Michael Modro


 

[GoShen]

That particular server may have desynced. Have we attempted a reboot of that problem server? I would also check the services and check event log to see if they have been stopping from the Windows Event Viewer.

Edited September 8 by GoShen
More Info

[Michael Modro]

On 9/8/2024 at 9:01 AM, GoShen said:

That particular server may have desynced. Have we attempted a reboot of that problem server? I would also check the services and check event log to see if they have been stopping from the Windows Event Viewer.

Thanks for your feedback. 

A reboot helps. After a reboot the problem connection server runs normaly again.

If the failure appears all services are  running and nothing is logged in the event log.
But if there would be a logged event, the NSX advanced load balancer cannot triggered for that.
The favicon check is then still healty.

we need a better check over REST API as the favicon check, to switch over the new sessions to a running connection server without failure.

I am wondering that not more admins have the same problem. 
If the health check don't work better the users get no connection, because of the NSX advanced load balancer didn't knwo this and not switched to other servers in the server pool.

 

[Sean Massey-1]

There is definitely something going on with that connection server that is causing issues without taking the service offline.  Normally, when Horizon stops responding to requests, the entire web server component is offline, so favicon.ico fails.

Have you opened a support ticket to investigate the issue that is causing the server to fail?  I would start there because this is not normal behavior.  

For now, I would recommend disabling the bad connection server inside your AVI pool so that no connection attempts are sent to it and open a ticket to investigate.