Double-check the services on the inactive node to make sure the System Attendant service hasn't been set to disabled and is still set to Local System. You've checked both Application and System Events?
So it's not being used in a Frontend - Backend scenario? The cluster itself is configured to serve up the actual OWA site? Do not make any port changes through IIS. Make them all through System Manager on the HTTP Virtual Server. Whenever the IIS Metadirectory information refreshes it will reset IIS to what Exchange "knows". If you have trouble setting it, you must clear out the port 80 that's already filled in before it will let you add an SSL port for the IP Address. From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Richard, Pat Sent: Monday, August 20, 2007 1:41 PM To: Exchange Discussions Subject: SA won't start when failing over to passive node Greetings - I'm not the strongest cluster person, so this is baffling me. Client has a new, 2 node AP cluster w/ Exchange 2003. Installation originally ran fine, and failover and failback were flawless. However, the client noticed the other day that the cluster attempted to failover, and failed. It successfully failed back. When we manually attempt to failover, everything starts, except for the System Attendant. It fails, causing the cluster to failback. When that happens, everything comes up fine with a minor exception - the SSL port setting in IIS is now blank, causing OWA to not work. I change it to 443 and the world is a beautiful place. This is repeatable. In looking through the event logs, I don't see anything sticking out as to why it fails to failover. So - I'm appealing to you guys for some direction. What should I be looking for? -- Pat Richard, MCSE Senior Network Engineer Microsoft Exchange MVP ANALYSTS INTERNATIONAL <http://www.analysts.com/> <https://mvp.support.microsoft.com/profile=396B594E-87FF-44F1-B482-BC96227C4 26B>
<<image001.gif>>
smime.p7s
Description: S/MIME cryptographic signature
