Hello,
I am creating a new Remote Desktop High Available cluster that contain the followin servers :
2 Gateway and web Access servers in an NLB node (srv-gw01 - srv-gw02)
2 Broker and RD Session host configured with the Broker High Availability (srv-tse01 - srv-tse02)
2 MS-SQL servers in a AlwaysOn cluster for the Broker HA Database (srv-db01 - srv-db02)
The High Availability of the cluster work as expected only if I disconnect my session before taking down a server as you can see with my tests :
Test 1 (working) :
1) Start a rdp session on the Gateway NLB floating IP
2) Work on the remote applications
3) Notice the name of the TS server I am connected on (srv-tse02 in this case)
4) Close my remote session (with CTRL + ALT + END)
5) Disconnect the network of a Gateway and the server srv-tse02
6) Start again a rdp session
--> This work fine I work now on srv-tse01
Test 2 (not working) :
1) Start a rdp session on the Gateway NLB floating IP
2) Work on the remote applications
3) Notice the name of the TS server I am connected on (srv-tse01 in this case)
5) Disconnect the network of a Gateway and the server srv-tse01
--> The remote application stuck (normal), tell me "Reconnecting" but never reconnect. If I try to start a new session it still not work, seems to try to connect me on srv-tse01 (that is down) as I see in the event log :
Log Name : Microsoft-Windows-TerminalServices-SessionBroker/Operational Message : Service Broker for remote desktop connections has successfully processed the connection request for user DOMAIN\myuser. Redirection information: Target Name = SRV-TSE01 Target IP address = 192168111192 Netbios target = SRV-TSE01 Target = FQDN SRV-TSE01.domain.local Found disconnected session = 0x1.
It seems that if a rdp session still exist on a TSE server it try to connect me on that server even if it's down. But this is not the way it should work (in my opinion).
Has anyone encountered that problem before ? Any suggestions will be realy apprecied :)
Best regards
Gabriel