Hi,
We have a system where usually around 300+ users are logged in simultaneously. These users are working on +-20 Windows Server 2008 R2 Terminal Servers. These servers are in a cluster. We have a physical load balancer. We have a Windows Server 2008 R2 Connection Broker. We also have serveral Windows Server 2003 servers, connecting to the same domain controllers, but to a different Farm on the Connection Broker.
Sometimes, and for us it seems like random times, users start complaining they cannot log on to the servers anymore. The server seems to be hanging on the User Profile Service. Users get the message "Please wait for the User Profile Service". This message will stay untill you give the server a hard reset (power off - power on). If we connect to the service from another computer in the domain, we can see which services are started. We can't see any problems, everything seems to be working fine. We just can't log on to the server anymore, also not as console. This happens on all our 2008 servers in the farm. If it happens, it is usually to happen on more than 1 server (last week on 7 servers simultaneously). You can imagine how our customers react to this. Users who are allready logged in to the servers on which the problems occur do not notice anything, untill they try to log out. Then they also get the message. If we shutdown this machine (some work on VMware, some are physical servers) the hard way (power off / power on) the server can work good for several weeks. Then, out of the blue, the problem might occur again.
We have tried many hotfixes and Windows Updates, we have searched in the DNS system, and disabled the second NIC, but the problem still occurs. We do have mandatory profiles, which are being loaded by the software "Flex Profiles".
I hope somebody here has some good ideas.
Thanks in advantage,
Roy