Just a quick post about this technote: http://technet.microsoft.com/en-us/library/cc671176.aspx
It claims this bug was fixed in rollup 4, but after struggling with a server with rollup 7 installed for many hours today, I can confirm that this bug is not fixed, and you do need to follow the procedure in the technote.
The bug concerns Outlook Anywhere (what used to be called RPC over HTTP). If the Exchange 2007 server is installed on Windows 2008 server your clients can’t connect until you follow the technote and then reboot the Exchange server.
Unfortunately, after implementing this “fix” Outlook Anywhere was working – but the SCR replication I had set up between the servers was broken! The replication status was “Disabled”. I tried everything to get it started again but was getting a bunch of new errors about having used a “simple server name” instead of the FQDN – despite using exactly the same powershell commands that had worked before.
Finally I backed out the hosts file change from the above technote and I’m back where I was before – SCR working but Outlook Anywhere broken.