[EMAIL PROTECTED] wrote: > On Mon, 29 Jan 2007, Christopher D. Clausen wrote: >> [EMAIL PROTECTED] wrote: >> >>> I'm moving the server to a new cluster of RHE hosts that use virtual >>> interfaces (eg. eth0:1) to allow for failover to a new host while >>> still maintaining the original IP address. On this new system I'm >>> getting the following error when I run sshd in debug (-ddd) mode: >>> >>> Wrong principal in request >>> >>> I have 2 IP addresses and 2 hostnames associated with the 2 >>> interfaces (one of them a virtual interface) on my workstation: >>> >>> interface hostname ip >>> ----------------------------------------- >>> eth0 gort.home.org 192.168.0.2 >>> eth0:1 cvs.home.org 192.168.0.200 >> >> Can you simply fail-over using the same IP on both interfaces? (I >> believe there is a bonding module in Linux that can do this.) > > The point of the virt interface is so it can be moved to a different > host. If the virt interface has the same IP as the real interface, > then it couldn't be moved to another host. In other words, the > "fail-over" is to fail over to a completely separate host, not a > separate interface on the same host.
Sorry, I think I'm missing something... These are NOT Kerberos KDCs are they? You are trying to have a clustered service that uses Kerberos for SSH? And can essentially be treated a multi-homed system? Do you have proper A and PTR records for both names? What does your /etc/hosts file look like? What does hostname -f return on your system? <<CDC ________________________________________________ Kerberos mailing list Kerberos@mit.edu https://mailman.mit.edu/mailman/listinfo/kerberos